diff --git a/log_rank0_1742725966.txt b/log_rank0_1742725966.txt new file mode 100644 index 0000000000000000000000000000000000000000..1d959f745d508a1bbc1d768fccd0e9e07e7eb819 --- /dev/null +++ b/log_rank0_1742725966.txt @@ -0,0 +1,1775 @@ +[2025-03-23 11:32:46 root] (main.py 352): INFO Namespace(net=None, model='/home/nianhui.guo/models/deepseek/models--deepseek-ai--DeepSeek-R1/snapshots/a157fa3d494497a54586a333a23df6c2143e7697/', cache_dir='./cache', output_dir='./log/deepseek-r1-w4a8g128-sample128-seqlen4096-search-layer-mix-gpu-batch4-hadmad-stage1-search/', save_dir='./log/deepseek-r1-w4a8g128-sample128-seqlen4096-search-layer-mix-gpu-batch4-hadmad-stage1-search/', resume=None, real_quant=False, calib_dataset='wikitext2', nsamples=4, batch_size=1, mini_batch_size=1, seed=2, epochs=1, reverse_epochs=3, limit=-1, deactive_amp=False, clip_grad=1.0, seq_length=2048, gradient_checkpoint=False, blocks=0, reverse_block_qat=False, top_calibration=True, top_calibration_blocks=1, recurrent_calibration_iterations=1, bidirection_qat=True, tasks='', eval_ppl=True, num_fewshot=0, eval_seq_length=2048, wbits=4.2, abits=16, symmetric=False, group_size=128, act_group_size=-1, qat=False, shift=True, optimizer='deepspeed', wd=0, norm_lr=1e-05, scale_lr=1e-05, one_bit_weight_lr=1e-05, two_bit_weight_lr=1e-06, four_bit_weight_lr=5e-06, shift_lr=1e-05, post_lr=1e-05, channel_scaler_lr=1e-06, prepare_lr=5e-06, fp_lr=1e-06, aug_loss=False, layerwise_loss=False, attention_loss=False, kl_loss=False, lm_head_loss=False, search=True, measurement=False, strategy=False, measure_dir='./measure/deepseek-r1-sample128-gpu-test/', strategy_dir='./strategy/deepseek-r1-sample128-gpu/', nonuniform_layout=False, residual_ratio=0.1, base_rank_ratio=0.01, exl_v2_pack=False, mix_precision_layout='layer_mix', rotation=False, online_rotation=False, rotate_mode='hadamard', lora_adapter=False, attn_implementation='sdpa', local_rank=-1, master_port=25901) +[2025-03-23 11:37:17 root] (main.py 383): INFO === start quantization === +[2025-03-23 11:37:34 root] (omniquant.py 130): INFO Starting ... +[2025-03-23 11:37:43 root] (measure_moe.py 464): INFO === Start search layer 0 mix precision layout for DeepSeek MoE === +[2025-03-23 11:37:51 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 0 +[2025-03-23 11:38:18 root] (measure_moe.py 516): INFO Layer 0 measurement completed in 35.36 seconds +[2025-03-23 11:38:18 root] (measure_moe.py 464): INFO === Start search layer 1 mix precision layout for DeepSeek MoE === +[2025-03-23 11:38:25 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 1 +[2025-03-23 11:38:52 root] (measure_moe.py 516): INFO Layer 1 measurement completed in 34.51 seconds +[2025-03-23 11:38:53 root] (measure_moe.py 464): INFO === Start search layer 2 mix precision layout for DeepSeek MoE === +[2025-03-23 11:39:01 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 2 +[2025-03-23 11:39:27 root] (measure_moe.py 516): INFO Layer 2 measurement completed in 34.50 seconds +[2025-03-23 11:39:27 root] (measure_moe.py 464): INFO === Start search layer 3 mix precision layout for DeepSeek MoE === +[2025-03-23 11:41:38 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 3 +[2025-03-23 11:46:05 root] (measure_moe.py 516): INFO Layer 3 measurement completed in 398.33 seconds +[2025-03-23 11:46:05 root] (measure_moe.py 464): INFO === Start search layer 4 mix precision layout for DeepSeek MoE === +[2025-03-23 11:48:12 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 4 +[2025-03-23 11:52:13 root] (measure_moe.py 516): INFO Layer 4 measurement completed in 367.11 seconds +[2025-03-23 11:52:13 root] (measure_moe.py 464): INFO === Start search layer 5 mix precision layout for DeepSeek MoE === +[2025-03-23 11:54:22 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 5 +[2025-03-23 11:58:35 root] (measure_moe.py 516): INFO Layer 5 measurement completed in 382.25 seconds +[2025-03-23 11:58:35 root] (measure_moe.py 464): INFO === Start search layer 6 mix precision layout for DeepSeek MoE === +[2025-03-23 12:00:53 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 6 +[2025-03-23 12:05:06 root] (measure_moe.py 516): INFO Layer 6 measurement completed in 391.42 seconds +[2025-03-23 12:05:06 root] (measure_moe.py 464): INFO === Start search layer 7 mix precision layout for DeepSeek MoE === +[2025-03-23 12:07:24 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 7 +[2025-03-23 12:11:31 root] (measure_moe.py 516): INFO Layer 7 measurement completed in 384.18 seconds +[2025-03-23 12:11:31 root] (measure_moe.py 464): INFO === Start search layer 8 mix precision layout for DeepSeek MoE === +[2025-03-23 12:14:00 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 8 +[2025-03-23 12:17:57 root] (measure_moe.py 516): INFO Layer 8 measurement completed in 386.89 seconds +[2025-03-23 12:17:57 root] (measure_moe.py 464): INFO === Start search layer 9 mix precision layout for DeepSeek MoE === +[2025-03-23 12:20:12 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 9 +[2025-03-23 12:24:13 root] (measure_moe.py 516): INFO Layer 9 measurement completed in 375.30 seconds +[2025-03-23 12:24:13 root] (measure_moe.py 464): INFO === Start search layer 10 mix precision layout for DeepSeek MoE === +[2025-03-23 12:26:31 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 10 +[2025-03-23 12:30:24 root] (measure_moe.py 516): INFO Layer 10 measurement completed in 370.80 seconds +[2025-03-23 12:30:24 root] (measure_moe.py 464): INFO === Start search layer 11 mix precision layout for DeepSeek MoE === +[2025-03-23 12:33:46 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 11 +[2025-03-23 12:37:23 root] (measure_moe.py 516): INFO Layer 11 measurement completed in 419.79 seconds +[2025-03-23 12:37:24 root] (measure_moe.py 464): INFO === Start search layer 12 mix precision layout for DeepSeek MoE === +[2025-03-23 12:39:31 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 12 +[2025-03-23 12:43:18 root] (measure_moe.py 516): INFO Layer 12 measurement completed in 354.95 seconds +[2025-03-23 12:43:19 root] (measure_moe.py 464): INFO === Start search layer 13 mix precision layout for DeepSeek MoE === +[2025-03-23 12:45:29 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 13 +[2025-03-23 12:49:37 root] (measure_moe.py 516): INFO Layer 13 measurement completed in 378.38 seconds +[2025-03-23 12:49:37 root] (measure_moe.py 464): INFO === Start search layer 14 mix precision layout for DeepSeek MoE === +[2025-03-23 12:51:50 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 14 +[2025-03-23 12:55:38 root] (measure_moe.py 516): INFO Layer 14 measurement completed in 361.43 seconds +[2025-03-23 12:55:38 root] (measure_moe.py 464): INFO === Start search layer 15 mix precision layout for DeepSeek MoE === +[2025-03-23 12:57:50 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 15 +[2025-03-23 13:02:00 root] (measure_moe.py 516): INFO Layer 15 measurement completed in 381.92 seconds +[2025-03-23 13:02:00 root] (measure_moe.py 464): INFO === Start search layer 16 mix precision layout for DeepSeek MoE === +[2025-03-23 13:04:15 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 16 +[2025-03-23 13:08:26 root] (measure_moe.py 516): INFO Layer 16 measurement completed in 385.33 seconds +[2025-03-23 13:08:26 root] (measure_moe.py 464): INFO === Start search layer 17 mix precision layout for DeepSeek MoE === +[2025-03-23 13:11:11 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 17 +[2025-03-23 13:15:12 root] (measure_moe.py 516): INFO Layer 17 measurement completed in 406.10 seconds +[2025-03-23 13:15:12 root] (measure_moe.py 464): INFO === Start search layer 18 mix precision layout for DeepSeek MoE === +[2025-03-23 13:17:28 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 18 +[2025-03-23 13:21:15 root] (measure_moe.py 516): INFO Layer 18 measurement completed in 362.74 seconds +[2025-03-23 13:21:15 root] (measure_moe.py 464): INFO === Start search layer 19 mix precision layout for DeepSeek MoE === +[2025-03-23 13:23:28 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 19 +[2025-03-23 13:27:36 root] (measure_moe.py 516): INFO Layer 19 measurement completed in 381.69 seconds +[2025-03-23 13:27:36 root] (measure_moe.py 464): INFO === Start search layer 20 mix precision layout for DeepSeek MoE === +[2025-03-23 13:29:52 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 20 +[2025-03-23 13:33:43 root] (measure_moe.py 516): INFO Layer 20 measurement completed in 366.88 seconds +[2025-03-23 13:33:43 root] (measure_moe.py 464): INFO === Start search layer 21 mix precision layout for DeepSeek MoE === +[2025-03-23 13:36:32 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 21 +[2025-03-23 13:40:07 root] (measure_moe.py 516): INFO Layer 21 measurement completed in 383.26 seconds +[2025-03-23 13:40:07 root] (measure_moe.py 464): INFO === Start search layer 22 mix precision layout for DeepSeek MoE === +[2025-03-23 13:42:28 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 22 +[2025-03-23 13:46:23 root] (measure_moe.py 516): INFO Layer 22 measurement completed in 376.62 seconds +[2025-03-23 13:46:23 root] (measure_moe.py 464): INFO === Start search layer 23 mix precision layout for DeepSeek MoE === +[2025-03-23 13:48:43 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 23 +[2025-03-23 13:52:40 root] (measure_moe.py 516): INFO Layer 23 measurement completed in 376.91 seconds +[2025-03-23 13:52:40 root] (measure_moe.py 464): INFO === Start search layer 24 mix precision layout for DeepSeek MoE === +[2025-03-23 13:55:24 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 24 +[2025-03-23 13:59:11 root] (measure_moe.py 516): INFO Layer 24 measurement completed in 391.23 seconds +[2025-03-23 13:59:11 root] (measure_moe.py 464): INFO === Start search layer 25 mix precision layout for DeepSeek MoE === +[2025-03-23 14:01:24 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 25 +[2025-03-23 14:05:16 root] (measure_moe.py 516): INFO Layer 25 measurement completed in 364.56 seconds +[2025-03-23 14:05:16 root] (measure_moe.py 464): INFO === Start search layer 26 mix precision layout for DeepSeek MoE === +[2025-03-23 14:07:32 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 26 +[2025-03-23 14:11:22 root] (measure_moe.py 516): INFO Layer 26 measurement completed in 365.90 seconds +[2025-03-23 14:11:22 root] (measure_moe.py 464): INFO === Start search layer 27 mix precision layout for DeepSeek MoE === +[2025-03-23 14:13:35 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 27 +[2025-03-23 14:17:21 root] (measure_moe.py 516): INFO Layer 27 measurement completed in 359.11 seconds +[2025-03-23 14:17:21 root] (measure_moe.py 464): INFO === Start search layer 28 mix precision layout for DeepSeek MoE === +[2025-03-23 14:19:39 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 28 +[2025-03-23 14:23:24 root] (measure_moe.py 516): INFO Layer 28 measurement completed in 363.15 seconds +[2025-03-23 14:23:24 root] (measure_moe.py 464): INFO === Start search layer 29 mix precision layout for DeepSeek MoE === +[2025-03-23 14:25:42 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 29 +[2025-03-23 14:29:29 root] (measure_moe.py 516): INFO Layer 29 measurement completed in 364.97 seconds +[2025-03-23 14:29:29 root] (measure_moe.py 464): INFO === Start search layer 30 mix precision layout for DeepSeek MoE === +[2025-03-23 14:31:45 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 30 +[2025-03-23 14:35:29 root] (measure_moe.py 516): INFO Layer 30 measurement completed in 360.02 seconds +[2025-03-23 14:35:29 root] (measure_moe.py 464): INFO === Start search layer 31 mix precision layout for DeepSeek MoE === +[2025-03-23 14:37:47 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 31 +[2025-03-23 14:41:29 root] (measure_moe.py 516): INFO Layer 31 measurement completed in 359.44 seconds +[2025-03-23 14:41:29 root] (measure_moe.py 464): INFO === Start search layer 32 mix precision layout for DeepSeek MoE === +[2025-03-23 14:43:44 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 32 +[2025-03-23 14:47:32 root] (measure_moe.py 516): INFO Layer 32 measurement completed in 363.36 seconds +[2025-03-23 14:47:32 root] (measure_moe.py 464): INFO === Start search layer 33 mix precision layout for DeepSeek MoE === +[2025-03-23 14:49:50 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 33 +[2025-03-23 14:53:33 root] (measure_moe.py 516): INFO Layer 33 measurement completed in 360.35 seconds +[2025-03-23 14:53:33 root] (measure_moe.py 464): INFO === Start search layer 34 mix precision layout for DeepSeek MoE === +[2025-03-23 14:55:49 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 34 +[2025-03-23 14:59:41 root] (measure_moe.py 516): INFO Layer 34 measurement completed in 367.92 seconds +[2025-03-23 14:59:41 root] (measure_moe.py 464): INFO === Start search layer 35 mix precision layout for DeepSeek MoE === +[2025-03-23 15:01:57 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 35 +[2025-03-23 15:05:40 root] (measure_moe.py 516): INFO Layer 35 measurement completed in 359.17 seconds +[2025-03-23 15:05:40 root] (measure_moe.py 464): INFO === Start search layer 36 mix precision layout for DeepSeek MoE === +[2025-03-23 15:07:56 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 36 +[2025-03-23 15:11:40 root] (measure_moe.py 516): INFO Layer 36 measurement completed in 360.50 seconds +[2025-03-23 15:11:40 root] (measure_moe.py 464): INFO === Start search layer 37 mix precision layout for DeepSeek MoE === +[2025-03-23 15:13:55 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 37 +[2025-03-23 15:17:40 root] (measure_moe.py 516): INFO Layer 37 measurement completed in 359.79 seconds +[2025-03-23 15:17:40 root] (measure_moe.py 464): INFO === Start search layer 38 mix precision layout for DeepSeek MoE === +[2025-03-23 15:19:55 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 38 +[2025-03-23 15:23:34 root] (measure_moe.py 516): INFO Layer 38 measurement completed in 354.02 seconds +[2025-03-23 15:23:34 root] (measure_moe.py 464): INFO === Start search layer 39 mix precision layout for DeepSeek MoE === +[2025-03-23 15:25:51 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 39 +[2025-03-23 15:29:32 root] (measure_moe.py 516): INFO Layer 39 measurement completed in 357.52 seconds +[2025-03-23 15:29:32 root] (measure_moe.py 464): INFO === Start search layer 40 mix precision layout for DeepSeek MoE === +[2025-03-23 15:31:49 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 40 +[2025-03-23 15:35:31 root] (measure_moe.py 516): INFO Layer 40 measurement completed in 359.38 seconds +[2025-03-23 15:35:31 root] (measure_moe.py 464): INFO === Start search layer 41 mix precision layout for DeepSeek MoE === +[2025-03-23 15:37:48 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 41 +[2025-03-23 15:41:32 root] (measure_moe.py 516): INFO Layer 41 measurement completed in 360.33 seconds +[2025-03-23 15:41:32 root] (measure_moe.py 464): INFO === Start search layer 42 mix precision layout for DeepSeek MoE === +[2025-03-23 15:43:50 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 42 +[2025-03-23 15:47:33 root] (measure_moe.py 516): INFO Layer 42 measurement completed in 361.41 seconds +[2025-03-23 15:47:33 root] (measure_moe.py 464): INFO === Start search layer 43 mix precision layout for DeepSeek MoE === +[2025-03-23 15:49:58 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 43 +[2025-03-23 15:53:42 root] (measure_moe.py 516): INFO Layer 43 measurement completed in 368.55 seconds +[2025-03-23 15:53:42 root] (measure_moe.py 464): INFO === Start search layer 44 mix precision layout for DeepSeek MoE === +[2025-03-23 15:55:59 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 44 +[2025-03-23 15:59:48 root] (measure_moe.py 516): INFO Layer 44 measurement completed in 365.82 seconds +[2025-03-23 15:59:48 root] (measure_moe.py 464): INFO === Start search layer 45 mix precision layout for DeepSeek MoE === +[2025-03-23 16:02:05 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 45 +[2025-03-23 16:05:51 root] (measure_moe.py 516): INFO Layer 45 measurement completed in 363.00 seconds +[2025-03-23 16:05:51 root] (measure_moe.py 464): INFO === Start search layer 46 mix precision layout for DeepSeek MoE === +[2025-03-23 16:08:05 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 46 +[2025-03-23 16:11:47 root] (measure_moe.py 516): INFO Layer 46 measurement completed in 356.26 seconds +[2025-03-23 16:11:47 root] (measure_moe.py 464): INFO === Start search layer 47 mix precision layout for DeepSeek MoE === +[2025-03-23 16:14:04 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 47 +[2025-03-23 16:17:46 root] (measure_moe.py 516): INFO Layer 47 measurement completed in 359.11 seconds +[2025-03-23 16:17:46 root] (measure_moe.py 464): INFO === Start search layer 48 mix precision layout for DeepSeek MoE === +[2025-03-23 16:20:02 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 48 +[2025-03-23 16:23:44 root] (measure_moe.py 516): INFO Layer 48 measurement completed in 357.67 seconds +[2025-03-23 16:23:44 root] (measure_moe.py 464): INFO === Start search layer 49 mix precision layout for DeepSeek MoE === +[2025-03-23 16:25:59 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 49 +[2025-03-23 16:29:40 root] (measure_moe.py 516): INFO Layer 49 measurement completed in 356.26 seconds +[2025-03-23 16:29:40 root] (measure_moe.py 464): INFO === Start search layer 50 mix precision layout for DeepSeek MoE === +[2025-03-23 16:31:58 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 50 +[2025-03-23 16:35:38 root] (measure_moe.py 516): INFO Layer 50 measurement completed in 357.87 seconds +[2025-03-23 16:35:38 root] (measure_moe.py 464): INFO === Start search layer 51 mix precision layout for DeepSeek MoE === +[2025-03-23 16:37:57 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 51 +[2025-03-23 16:41:39 root] (measure_moe.py 516): INFO Layer 51 measurement completed in 360.55 seconds +[2025-03-23 16:41:39 root] (measure_moe.py 464): INFO === Start search layer 52 mix precision layout for DeepSeek MoE === +[2025-03-23 16:43:56 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 52 +[2025-03-23 16:47:32 root] (measure_moe.py 516): INFO Layer 52 measurement completed in 353.56 seconds +[2025-03-23 16:47:32 root] (measure_moe.py 464): INFO === Start search layer 53 mix precision layout for DeepSeek MoE === +[2025-03-23 16:49:51 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 53 +[2025-03-23 16:53:19 root] (measure_moe.py 516): INFO Layer 53 measurement completed in 346.55 seconds +[2025-03-23 16:53:19 root] (measure_moe.py 464): INFO === Start search layer 54 mix precision layout for DeepSeek MoE === +[2025-03-23 16:55:37 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 54 +[2025-03-23 16:59:25 root] (measure_moe.py 516): INFO Layer 54 measurement completed in 366.55 seconds +[2025-03-23 16:59:25 root] (measure_moe.py 464): INFO === Start search layer 55 mix precision layout for DeepSeek MoE === +[2025-03-23 17:01:42 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 55 +[2025-03-23 17:05:09 root] (measure_moe.py 516): INFO Layer 55 measurement completed in 343.86 seconds +[2025-03-23 17:05:09 root] (measure_moe.py 464): INFO === Start search layer 56 mix precision layout for DeepSeek MoE === +[2025-03-23 17:11:17 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 56 +[2025-03-23 17:14:43 root] (measure_moe.py 516): INFO Layer 56 measurement completed in 573.69 seconds +[2025-03-23 17:14:43 root] (measure_moe.py 464): INFO === Start search layer 57 mix precision layout for DeepSeek MoE === +[2025-03-23 17:18:01 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 57 +[2025-03-23 17:21:54 root] (measure_moe.py 516): INFO Layer 57 measurement completed in 430.55 seconds +[2025-03-23 17:21:54 root] (measure_moe.py 464): INFO === Start search layer 58 mix precision layout for DeepSeek MoE === +[2025-03-23 17:24:11 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 58 +[2025-03-23 17:28:12 root] (measure_moe.py 516): INFO Layer 58 measurement completed in 378.73 seconds +[2025-03-23 17:28:13 root] (measure_moe.py 464): INFO === Start search layer 59 mix precision layout for DeepSeek MoE === +[2025-03-23 17:30:31 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 59 +[2025-03-23 17:34:25 root] (measure_moe.py 516): INFO Layer 59 measurement completed in 372.46 seconds +[2025-03-23 17:34:25 root] (measure_moe.py 464): INFO === Start search layer 60 mix precision layout for DeepSeek MoE === +[2025-03-23 17:36:51 root] (measure_moe.py 509): INFO Using DeepSeek MoE specific measurement for layer 60 +[2025-03-23 17:40:41 root] (measure_moe.py 516): INFO Layer 60 measurement completed in 376.24 seconds +[2025-03-23 17:40:50 root] (omniquant.py 395): INFO === Start quantize layer 0:0 === +[2025-03-23 17:40:57 root] (omniquant.py 432): INFO model.layers.0.self_attn.q_a_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:40:57 root] (omniquant.py 432): INFO model.layers.0.self_attn.q_b_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:40:57 root] (omniquant.py 432): INFO model.layers.0.self_attn.kv_a_proj_with_mqa: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:40:57 root] (omniquant.py 432): INFO model.layers.0.self_attn.kv_b_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:40:57 root] (omniquant.py 432): INFO model.layers.0.self_attn.o_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:40:57 root] (omniquant.py 432): INFO model.layers.0.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:40:57 root] (omniquant.py 432): INFO model.layers.0.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:40:57 root] (omniquant.py 432): INFO model.layers.0.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:04 root] (omniquant.py 804): INFO layer 0 iter 0 loss:0.000521926733199507 fp_loss:0 norm:0.0 lr:1e-05 max GPU memory_allocated 50653.25537109375 +[2025-03-23 17:41:07 root] (omniquant.py 804): INFO layer 0 iter 1 loss:0.00014829730207566172 fp_loss:0 norm:0.0 lr:1e-05 max GPU memory_allocated 50653.25537109375 +[2025-03-23 17:41:11 root] (omniquant.py 395): INFO === Start quantize layer 1:1 === +[2025-03-23 17:41:18 root] (omniquant.py 432): INFO model.layers.1.self_attn.q_a_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:18 root] (omniquant.py 432): INFO model.layers.1.self_attn.q_b_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:18 root] (omniquant.py 432): INFO model.layers.1.self_attn.kv_a_proj_with_mqa: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:18 root] (omniquant.py 432): INFO model.layers.1.self_attn.kv_b_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:19 root] (omniquant.py 432): INFO model.layers.1.self_attn.o_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:19 root] (omniquant.py 432): INFO model.layers.1.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:19 root] (omniquant.py 432): INFO model.layers.1.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:19 root] (omniquant.py 432): INFO model.layers.1.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:26 root] (omniquant.py 804): INFO layer 1 iter 0 loss:0.0001148075534729287 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 50653.25537109375 +[2025-03-23 17:41:30 root] (omniquant.py 804): INFO layer 1 iter 1 loss:6.701215170323849e-05 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 50653.25537109375 +[2025-03-23 17:41:35 root] (omniquant.py 395): INFO === Start quantize layer 2:2 === +[2025-03-23 17:41:43 root] (omniquant.py 432): INFO model.layers.2.self_attn.q_a_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:43 root] (omniquant.py 432): INFO model.layers.2.self_attn.q_b_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:43 root] (omniquant.py 432): INFO model.layers.2.self_attn.kv_a_proj_with_mqa: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:43 root] (omniquant.py 432): INFO model.layers.2.self_attn.kv_b_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:43 root] (omniquant.py 432): INFO model.layers.2.self_attn.o_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:43 root] (omniquant.py 432): INFO model.layers.2.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:44 root] (omniquant.py 432): INFO model.layers.2.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:44 root] (omniquant.py 432): INFO model.layers.2.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:41:48 root] (omniquant.py 804): INFO layer 2 iter 0 loss:0.0002485747681930661 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 50653.25537109375 +[2025-03-23 17:41:50 root] (omniquant.py 804): INFO layer 2 iter 1 loss:0.00013705329911317676 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 50653.25537109375 +[2025-03-23 17:41:58 root] (omniquant.py 395): INFO === Start quantize layer 3:3 === +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.self_attn.q_a_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.self_attn.q_b_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.self_attn.kv_a_proj_with_mqa: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.self_attn.kv_b_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.self_attn.o_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.0.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.0.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.0.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.1.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.1.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.1.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.2.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.2.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.2.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.3.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.3.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.3.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.4.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.4.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.4.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:04 root] (omniquant.py 432): INFO model.layers.3.experts.5.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.5.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.5.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.6.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.6.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.6.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.7.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.7.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.7.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.8.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.8.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.8.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.9.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.9.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.9.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.10.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.10.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.10.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.11.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.11.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.11.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.12.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.12.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.12.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.13.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.13.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.13.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.14.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.14.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.14.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.15.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.15.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.15.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.16.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.16.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.16.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.17.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.17.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.17.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.18.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.18.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.18.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.19.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.19.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.19.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.20.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.20.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.20.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.21.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.21.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.21.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.22.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.22.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:05 root] (omniquant.py 432): INFO model.layers.3.experts.22.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.23.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.23.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.23.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.24.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.24.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.24.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.25.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.25.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.25.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.26.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.26.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.26.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.27.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.27.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.27.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.28.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.28.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.28.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.29.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.29.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.29.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.30.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.30.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.30.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.31.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.31.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.31.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.32.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.32.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.32.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.33.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.33.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.33.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.34.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.34.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.34.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.35.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.35.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.35.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.36.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.36.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.36.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.37.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.37.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.37.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.38.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.38.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.38.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.39.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.39.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.39.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.40.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:06 root] (omniquant.py 432): INFO model.layers.3.experts.40.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.40.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.41.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.41.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.41.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.42.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.42.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.42.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.43.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.43.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.43.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.44.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.44.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.44.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.45.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.45.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.45.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.46.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.46.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.46.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.47.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.47.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.47.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.48.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.48.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.48.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.49.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.49.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.49.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.50.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.50.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.50.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.51.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.51.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.51.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.52.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.52.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.52.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.53.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.53.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.53.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.54.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.54.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.54.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.55.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.55.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.55.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.56.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.56.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.56.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.57.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.57.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.57.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:07 root] (omniquant.py 432): INFO model.layers.3.experts.58.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.58.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.58.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.59.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.59.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.59.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.60.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.60.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.60.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.61.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.61.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.61.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.62.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.62.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.62.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.63.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.63.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.63.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.64.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.64.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.64.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.65.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.65.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.65.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.66.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.66.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.66.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.67.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.67.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.67.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.68.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.68.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.68.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.69.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.69.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.69.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.70.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.70.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.70.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.71.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.71.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.71.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.72.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.72.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.72.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.73.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.73.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.73.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.74.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.74.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.74.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.75.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.75.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:08 root] (omniquant.py 432): INFO model.layers.3.experts.75.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.76.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.76.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.76.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.77.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.77.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.77.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.78.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.78.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.78.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.79.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.79.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.79.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.80.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.80.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.80.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.81.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.81.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.81.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.82.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.82.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.82.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.83.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.83.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.83.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.84.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.84.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.84.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.85.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.85.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.85.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.86.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.86.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.86.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.87.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.87.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.87.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.88.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.88.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.88.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.89.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.89.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.89.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.90.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.90.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.90.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.91.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.91.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.91.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.92.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.92.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.92.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:09 root] (omniquant.py 432): INFO model.layers.3.experts.93.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.93.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.93.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.94.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.94.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.94.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.95.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.95.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.95.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.96.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.96.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.96.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.97.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.97.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.97.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.98.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.98.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.98.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.99.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.99.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.99.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.100.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.100.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.100.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.101.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.101.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.101.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.102.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.102.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.102.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.103.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.103.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.103.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.104.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.104.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.104.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.105.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.105.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.105.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.106.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.106.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.106.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.107.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.107.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.107.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.108.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.108.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.108.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.109.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.109.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.109.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.110.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.110.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:10 root] (omniquant.py 432): INFO model.layers.3.experts.110.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.111.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.111.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.111.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.112.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.112.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.112.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.113.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.113.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.113.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.114.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.114.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.114.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.115.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.115.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.115.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.116.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.116.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.116.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.117.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.117.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.117.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.118.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.118.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.118.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.119.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.119.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.119.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.120.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.120.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.120.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.121.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.121.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.121.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.122.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.122.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.122.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.123.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.123.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.123.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.124.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.124.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.124.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.125.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.125.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.125.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.126.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.126.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.126.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.127.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.127.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.127.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.128.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:11 root] (omniquant.py 432): INFO model.layers.3.experts.128.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.128.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.129.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.129.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.129.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.130.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.130.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.130.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.131.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.131.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.131.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.132.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.132.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.132.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.133.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.133.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.133.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.134.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.134.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.134.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.135.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.135.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.135.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.136.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.136.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.136.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.137.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.137.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.137.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.138.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.138.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.138.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.139.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.139.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.139.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.140.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.140.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.140.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.141.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.141.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.141.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.142.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.142.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.142.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.143.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.143.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.143.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.144.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.144.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.144.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.145.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.145.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.145.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:12 root] (omniquant.py 432): INFO model.layers.3.experts.146.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.146.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.146.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.147.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.147.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.147.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.148.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.148.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.148.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.149.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.149.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.149.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.150.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.150.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.150.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.151.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.151.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.151.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.152.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.152.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.152.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.153.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.153.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.153.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.154.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.154.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.154.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.155.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.155.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.155.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.156.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.156.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.156.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.157.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.157.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.157.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.158.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.158.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.158.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.159.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.159.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.159.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.160.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.160.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.160.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.161.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.161.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.161.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.162.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.162.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.162.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.163.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:13 root] (omniquant.py 432): INFO model.layers.3.experts.163.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.163.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.164.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.164.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.164.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.165.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.165.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.165.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.166.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.166.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.166.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.167.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.167.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.167.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.168.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.168.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.168.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.169.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.169.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.169.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.170.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.170.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.170.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.171.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.171.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.171.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.172.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.172.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.172.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.173.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.173.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.173.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.174.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.174.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.174.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.175.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.175.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.175.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.176.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.176.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.176.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.177.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.177.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.177.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.178.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.178.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.178.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.179.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.179.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.179.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.180.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.180.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.180.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.181.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.181.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:14 root] (omniquant.py 432): INFO model.layers.3.experts.181.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.182.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.182.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.182.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.183.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.183.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.183.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.184.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.184.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.184.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.185.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.185.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.185.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.186.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.186.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.186.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.187.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.187.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.187.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.188.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.188.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.188.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.189.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.189.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.189.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.190.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.190.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.190.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.191.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.191.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.191.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.192.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.192.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.192.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.193.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.193.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.193.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.194.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.194.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.194.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.195.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.195.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.195.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.196.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.196.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.196.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.197.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.197.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.197.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.198.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.198.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.198.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.199.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.199.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.199.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.200.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.200.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:15 root] (omniquant.py 432): INFO model.layers.3.experts.200.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.201.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.201.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.201.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.202.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.202.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.202.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.203.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.203.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.203.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.204.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.204.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.204.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.205.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.205.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.205.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.206.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.206.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.206.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.207.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.207.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.207.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.208.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.208.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.208.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.209.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.209.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.209.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.210.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.210.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.210.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.211.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.211.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.211.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.212.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.212.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.212.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.213.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.213.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.213.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.214.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.214.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.214.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.215.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.215.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.215.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.216.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.216.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.216.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.217.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.217.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.217.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.218.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:16 root] (omniquant.py 432): INFO model.layers.3.experts.218.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.218.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.219.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.219.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.219.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.220.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.220.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.220.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.221.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.221.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.221.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.222.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.222.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.222.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.223.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.223.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.223.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.224.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.224.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.224.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.225.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.225.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.225.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.226.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.226.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.226.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.227.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.227.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.227.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.228.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.228.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.228.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.229.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.229.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.229.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.230.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.230.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.230.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.231.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.231.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.231.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.232.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.232.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.232.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.233.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.233.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.233.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.234.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.234.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.234.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.235.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.235.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.235.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.236.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.236.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:17 root] (omniquant.py 432): INFO model.layers.3.experts.236.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.237.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.237.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.237.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.238.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.238.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.238.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.239.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.239.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.239.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.240.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.240.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.240.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.241.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.241.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.241.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.242.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.242.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.242.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.243.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.243.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.243.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.244.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.244.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.244.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.245.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.245.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.245.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.246.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.246.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.246.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.247.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.247.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.247.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.248.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.248.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.248.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.249.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.249.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.249.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.250.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.250.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.250.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.251.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.251.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.251.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.252.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.252.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.252.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.253.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.253.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.253.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.254.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.254.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:18 root] (omniquant.py 432): INFO model.layers.3.experts.254.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:19 root] (omniquant.py 432): INFO model.layers.3.experts.255.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:19 root] (omniquant.py 432): INFO model.layers.3.experts.255.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:19 root] (omniquant.py 432): INFO model.layers.3.experts.255.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:19 root] (omniquant.py 432): INFO model.layers.3.shared_experts.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:19 root] (omniquant.py 432): INFO model.layers.3.shared_experts.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:19 root] (omniquant.py 432): INFO model.layers.3.shared_experts.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:44:53 root] (omniquant.py 804): INFO layer 3 iter 0 loss:0.00021365127759054303 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 79615.8076171875 +[2025-03-23 17:45:17 root] (omniquant.py 804): INFO layer 3 iter 1 loss:0.00015112326946109533 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 79615.8076171875 +[2025-03-23 17:45:49 root] (omniquant.py 395): INFO === Start quantize layer 4:4 === +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.self_attn.q_a_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.self_attn.q_b_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.self_attn.kv_a_proj_with_mqa: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.self_attn.kv_b_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.self_attn.o_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.0.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.0.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.0.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.1.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.1.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.1.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.2.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.2.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.2.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.3.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.3.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.3.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.4.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.4.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.4.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.5.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.5.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.5.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.6.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.6.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.6.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.7.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.7.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.7.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.8.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.8.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.8.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.9.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.9.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:39 root] (omniquant.py 432): INFO model.layers.4.experts.9.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.10.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.10.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.10.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.11.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.11.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.11.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.12.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.12.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.12.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.13.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.13.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.13.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.14.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.14.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.14.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.15.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.15.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.15.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.16.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.16.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.16.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.17.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.17.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.17.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.18.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.18.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.18.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.19.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.19.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.19.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.20.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.20.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.20.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.21.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.21.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.21.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.22.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.22.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.22.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.23.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.23.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.23.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.24.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.24.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.24.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.25.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.25.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.25.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.26.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.26.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.26.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.27.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.27.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.27.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.28.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.28.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:40 root] (omniquant.py 432): INFO model.layers.4.experts.28.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.29.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.29.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.29.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.30.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.30.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.30.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.31.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.31.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.31.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.32.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.32.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.32.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.33.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.33.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.33.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.34.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.34.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.34.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.35.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.35.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.35.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.36.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.36.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.36.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.37.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.37.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.37.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.38.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.38.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.38.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.39.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.39.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.39.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.40.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.40.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.40.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.41.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.41.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.41.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.42.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.42.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.42.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.43.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.43.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.43.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.44.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.44.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.44.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.45.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.45.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.45.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.46.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.46.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.46.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.47.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.47.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.47.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:41 root] (omniquant.py 432): INFO model.layers.4.experts.48.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.48.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.48.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.49.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.49.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.49.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.50.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.50.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.50.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.51.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.51.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.51.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.52.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.52.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.52.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.53.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.53.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.53.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.54.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.54.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.54.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.55.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.55.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.55.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.56.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.56.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.56.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.57.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.57.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.57.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.58.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.58.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.58.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.59.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.59.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.59.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.60.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.60.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.60.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.61.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.61.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.61.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.62.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.62.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.62.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.63.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.63.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.63.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.64.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.64.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.64.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.65.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.65.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.65.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.66.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.66.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.66.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:42 root] (omniquant.py 432): INFO model.layers.4.experts.67.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.67.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.67.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.68.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.68.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.68.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.69.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.69.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.69.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.70.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.70.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.70.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.71.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.71.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.71.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.72.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.72.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.72.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.73.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.73.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.73.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.74.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.74.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.74.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.75.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.75.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.75.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.76.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.76.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.76.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.77.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.77.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.77.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.78.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.78.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.78.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.79.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.79.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.79.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.80.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.80.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.80.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.81.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.81.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.81.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.82.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.82.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.82.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.83.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.83.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.83.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.84.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.84.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.84.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.85.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:43 root] (omniquant.py 432): INFO model.layers.4.experts.85.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.85.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.86.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.86.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.86.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.87.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.87.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.87.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.88.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.88.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.88.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.89.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.89.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.89.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.90.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.90.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.90.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.91.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.91.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.91.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.92.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.92.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.92.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.93.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.93.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.93.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.94.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.94.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.94.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.95.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.95.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.95.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.96.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.96.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.96.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.97.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.97.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.97.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.98.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.98.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.98.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.99.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.99.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.99.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.100.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.100.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.100.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.101.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.101.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.101.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.102.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.102.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.102.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.103.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.103.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:44 root] (omniquant.py 432): INFO model.layers.4.experts.103.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.104.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.104.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.104.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.105.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.105.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.105.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.106.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.106.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.106.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.107.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.107.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.107.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.108.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.108.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.108.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.109.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.109.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.109.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.110.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.110.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.110.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.111.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.111.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.111.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.112.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.112.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.112.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.113.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.113.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.113.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.114.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.114.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.114.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.115.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.115.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.115.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.116.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.116.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.116.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.117.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.117.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.117.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.118.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.118.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.118.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.119.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.119.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.119.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.120.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.120.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.120.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.121.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.121.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:45 root] (omniquant.py 432): INFO model.layers.4.experts.121.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.122.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.122.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.122.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.123.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.123.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.123.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.124.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.124.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.124.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.125.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.125.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.125.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.126.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.126.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.126.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.127.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.127.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.127.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.128.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.128.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.128.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.129.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.129.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.129.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.130.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.130.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.130.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.131.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.131.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.131.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.132.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.132.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.132.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.133.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.133.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.133.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.134.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.134.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.134.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.135.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.135.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.135.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.136.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.136.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.136.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.137.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.137.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.137.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.138.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.138.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.138.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:46 root] (omniquant.py 432): INFO model.layers.4.experts.139.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.139.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.139.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.140.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.140.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.140.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.141.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.141.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.141.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.142.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.142.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.142.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.143.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.143.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.143.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.144.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.144.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.144.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.145.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.145.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.145.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.146.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.146.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.146.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.147.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.147.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.147.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.148.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.148.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.148.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.149.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.149.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.149.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.150.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.150.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.150.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.151.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.151.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.151.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.152.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.152.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.152.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.153.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.153.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.153.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.154.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.154.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.154.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.155.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.155.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.155.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.156.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.156.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.156.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:47 root] (omniquant.py 432): INFO model.layers.4.experts.157.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.157.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.157.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.158.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.158.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.158.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.159.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.159.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.159.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.160.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.160.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.160.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.161.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.161.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.161.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.162.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.162.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.162.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.163.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.163.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.163.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.164.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.164.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.164.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.165.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.165.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.165.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.166.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.166.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.166.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.167.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.167.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.167.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.168.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.168.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.168.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.169.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.169.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.169.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.170.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.170.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.170.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.171.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.171.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.171.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.172.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.172.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.172.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.173.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.173.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.173.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.174.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.174.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.174.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.175.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.175.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:48 root] (omniquant.py 432): INFO model.layers.4.experts.175.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.176.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.176.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.176.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.177.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.177.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.177.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.178.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.178.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.178.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.179.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.179.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.179.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.180.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.180.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.180.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.181.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.181.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.181.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.182.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.182.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.182.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.183.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.183.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.183.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.184.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.184.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.184.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.185.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.185.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.185.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.186.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.186.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.186.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.187.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.187.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.187.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.188.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.188.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.188.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.189.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.189.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.189.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.190.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.190.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.190.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.191.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.191.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.191.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.192.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.192.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.192.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:49 root] (omniquant.py 432): INFO model.layers.4.experts.193.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.193.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.193.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.194.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.194.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.194.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.195.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.195.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.195.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.196.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.196.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.196.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.197.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.197.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.197.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.198.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.198.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.198.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.199.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.199.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.199.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.200.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.200.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.200.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.201.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.201.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.201.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.202.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.202.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.202.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.203.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.203.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.203.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.204.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.204.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.204.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.205.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.205.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.205.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.206.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.206.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.206.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.207.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.207.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.207.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.208.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.208.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.208.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.209.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.209.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.209.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.210.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.210.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.210.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:50 root] (omniquant.py 432): INFO model.layers.4.experts.211.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.211.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.211.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.212.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.212.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.212.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.213.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.213.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.213.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.214.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.214.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.214.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.215.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.215.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.215.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.216.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.216.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.216.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.217.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.217.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.217.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.218.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.218.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.218.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.219.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.219.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.219.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.220.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.220.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.220.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.221.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.221.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.221.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.222.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.222.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.222.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.223.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.223.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.223.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.224.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.224.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.224.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.225.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.225.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.225.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.226.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.226.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.226.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.227.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.227.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.227.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.228.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.228.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:51 root] (omniquant.py 432): INFO model.layers.4.experts.228.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.229.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.229.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.229.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.230.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.230.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.230.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.231.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.231.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.231.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.232.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.232.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.232.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.233.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.233.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.233.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.234.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.234.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.234.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.235.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.235.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.235.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.236.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.236.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.236.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.237.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.237.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.237.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.238.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.238.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.238.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.239.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.239.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.239.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.240.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.240.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.240.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.241.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.241.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.241.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.242.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.242.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.242.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.243.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.243.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.243.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.244.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.244.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.244.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.245.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.245.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.245.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.246.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.246.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:52 root] (omniquant.py 432): INFO model.layers.4.experts.246.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.247.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.247.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.247.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.248.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.248.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.248.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.249.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.249.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.249.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.250.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.250.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.250.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.251.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.251.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.251.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.252.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.252.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.252.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.253.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.253.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.253.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.254.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.254.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.254.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.255.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.255.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.experts.255.mlp.up_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.shared_experts.mlp.gate_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.shared_experts.mlp.down_proj: QParams(32, [4], [1], 4, 32) +[2025-03-23 17:48:53 root] (omniquant.py 432): INFO model.layers.4.shared_experts.mlp.up_proj: QParams(32, [4], [1], 4, 32) diff --git a/log_rank0_1742749210.txt b/log_rank0_1742749210.txt new file mode 100644 index 0000000000000000000000000000000000000000..1fb80759a9f4c849459a05bdec1152d5f8c732da --- /dev/null +++ b/log_rank0_1742749210.txt @@ -0,0 +1,45219 @@ +[2025-03-23 18:00:10 root] (main.py 352): INFO Namespace(net=None, model='/home/nianhui.guo/models/deepseek/models--deepseek-ai--DeepSeek-R1/snapshots/a157fa3d494497a54586a333a23df6c2143e7697/', cache_dir='./cache', output_dir='./log/deepseek-r1-w4a8g128-sample128-seqlen4096-search-layer-mix-gpu-batch4-hadmad-stage1-search/', save_dir='./log/deepseek-r1-w4a8g128-sample128-seqlen4096-search-layer-mix-gpu-batch4-hadmad-stage1-search/', resume=None, real_quant=True, calib_dataset='wikitext2', nsamples=128, batch_size=1, mini_batch_size=1, seed=2, epochs=1, reverse_epochs=3, limit=-1, deactive_amp=False, clip_grad=1.0, seq_length=2048, gradient_checkpoint=False, blocks=0, reverse_block_qat=False, top_calibration=True, top_calibration_blocks=1, recurrent_calibration_iterations=1, bidirection_qat=True, tasks='', eval_ppl=True, num_fewshot=0, eval_seq_length=2048, wbits=4.2, abits=16, symmetric=False, group_size=128, act_group_size=-1, qat=False, shift=True, optimizer='deepspeed', wd=0, norm_lr=1e-05, scale_lr=1e-05, one_bit_weight_lr=1e-05, two_bit_weight_lr=1e-06, four_bit_weight_lr=5e-06, shift_lr=1e-05, post_lr=1e-05, channel_scaler_lr=1e-06, prepare_lr=5e-06, fp_lr=1e-06, aug_loss=False, layerwise_loss=False, attention_loss=False, kl_loss=False, lm_head_loss=False, search=True, measurement=True, strategy=False, measure_dir='./measure/deepseek-r1-sample128-gpu-test/', strategy_dir='./strategy/deepseek-r1-sample128-gpu/', nonuniform_layout=False, residual_ratio=0.1, base_rank_ratio=0.01, exl_v2_pack=False, mix_precision_layout='layer_mix', rotation=False, online_rotation=False, rotate_mode='hadamard', lora_adapter=False, attn_implementation='sdpa', local_rank=-1, master_port=25901) +[2025-03-23 18:04:52 root] (main.py 383): INFO === start quantization === +[2025-03-23 18:05:27 root] (omniquant.py 130): INFO Starting ... +[2025-03-23 18:05:59 root] (omniquant.py 395): INFO === Start quantize layer 0:0 === +[2025-03-23 18:06:05 root] (omniquant.py 432): INFO model.layers.0.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:06:05 root] (omniquant.py 432): INFO model.layers.0.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:06:05 root] (omniquant.py 432): INFO model.layers.0.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:06:05 root] (omniquant.py 432): INFO model.layers.0.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:06:05 root] (omniquant.py 432): INFO model.layers.0.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:06:05 root] (omniquant.py 432): INFO model.layers.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:06:05 root] (omniquant.py 432): INFO model.layers.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:06:05 root] (omniquant.py 432): INFO model.layers.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:08:43 root] (omniquant.py 804): INFO layer 0 iter 0 loss:3.900651790900156e-05 fp_loss:0 norm:0.0 lr:1e-05 max GPU memory_allocated 11274.6416015625 +[2025-03-23 18:11:19 root] (omniquant.py 804): INFO layer 0 iter 1 loss:5.632397915178444e-06 fp_loss:0 norm:0.0 lr:1e-05 max GPU memory_allocated 11274.642578125 +[2025-03-23 18:13:33 root] (omniquant.py 395): INFO === Start quantize layer 1:1 === +[2025-03-23 18:13:40 root] (omniquant.py 432): INFO model.layers.1.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:13:40 root] (omniquant.py 432): INFO model.layers.1.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:13:40 root] (omniquant.py 432): INFO model.layers.1.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:13:40 root] (omniquant.py 432): INFO model.layers.1.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:13:40 root] (omniquant.py 432): INFO model.layers.1.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:13:40 root] (omniquant.py 432): INFO model.layers.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:13:40 root] (omniquant.py 432): INFO model.layers.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:13:40 root] (omniquant.py 432): INFO model.layers.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:16:21 root] (omniquant.py 804): INFO layer 1 iter 0 loss:1.7975886294152588e-05 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 11274.642578125 +[2025-03-23 18:18:39 root] (omniquant.py 804): INFO layer 1 iter 1 loss:1.3209948519943282e-05 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 11274.642578125 +[2025-03-23 18:20:41 root] (omniquant.py 395): INFO === Start quantize layer 2:2 === +[2025-03-23 18:20:49 root] (omniquant.py 432): INFO model.layers.2.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:20:49 root] (omniquant.py 432): INFO model.layers.2.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:20:49 root] (omniquant.py 432): INFO model.layers.2.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:20:49 root] (omniquant.py 432): INFO model.layers.2.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:20:49 root] (omniquant.py 432): INFO model.layers.2.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:20:49 root] (omniquant.py 432): INFO model.layers.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:20:49 root] (omniquant.py 432): INFO model.layers.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:20:49 root] (omniquant.py 432): INFO model.layers.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:23:11 root] (omniquant.py 804): INFO layer 2 iter 0 loss:5.820677324663848e-05 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 11274.642578125 +[2025-03-23 18:25:31 root] (omniquant.py 804): INFO layer 2 iter 1 loss:2.00526301341597e-05 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 11274.642578125 +[2025-03-23 18:27:33 root] (omniquant.py 395): INFO === Start quantize layer 3:3 === +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:42 root] (omniquant.py 432): INFO model.layers.3.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:43 root] (omniquant.py 432): INFO model.layers.3.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:44 root] (omniquant.py 432): INFO model.layers.3.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:45 root] (omniquant.py 432): INFO model.layers.3.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:46 root] (omniquant.py 432): INFO model.layers.3.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:47 root] (omniquant.py 432): INFO model.layers.3.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:48 root] (omniquant.py 432): INFO model.layers.3.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:49 root] (omniquant.py 432): INFO model.layers.3.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:50 root] (omniquant.py 432): INFO model.layers.3.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:51 root] (omniquant.py 432): INFO model.layers.3.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:29:52 root] (omniquant.py 432): INFO model.layers.3.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 18:46:53 root] (omniquant.py 804): INFO layer 3 iter 0 loss:0.00011612522939685732 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-23 19:02:27 root] (omniquant.py 804): INFO layer 3 iter 1 loss:0.00010010242112912238 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-23 19:19:22 root] (omniquant.py 395): INFO === Start quantize layer 4:4 === +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:03 root] (omniquant.py 432): INFO model.layers.4.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:04 root] (omniquant.py 432): INFO model.layers.4.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:05 root] (omniquant.py 432): INFO model.layers.4.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:06 root] (omniquant.py 432): INFO model.layers.4.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:07 root] (omniquant.py 432): INFO model.layers.4.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:08 root] (omniquant.py 432): INFO model.layers.4.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:09 root] (omniquant.py 432): INFO model.layers.4.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:10 root] (omniquant.py 432): INFO model.layers.4.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:11 root] (omniquant.py 432): INFO model.layers.4.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:12 root] (omniquant.py 432): INFO model.layers.4.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:22:13 root] (omniquant.py 432): INFO model.layers.4.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 19:37:29 root] (omniquant.py 804): INFO layer 4 iter 0 loss:0.0003501121245790273 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-23 19:52:28 root] (omniquant.py 804): INFO layer 4 iter 1 loss:0.00025775376707315445 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-23 20:08:38 root] (omniquant.py 395): INFO === Start quantize layer 5:5 === +[2025-03-23 20:10:46 root] (omniquant.py 432): INFO model.layers.5.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:46 root] (omniquant.py 432): INFO model.layers.5.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:46 root] (omniquant.py 432): INFO model.layers.5.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:46 root] (omniquant.py 432): INFO model.layers.5.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:46 root] (omniquant.py 432): INFO model.layers.5.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:46 root] (omniquant.py 432): INFO model.layers.5.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:46 root] (omniquant.py 432): INFO model.layers.5.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:47 root] (omniquant.py 432): INFO model.layers.5.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:48 root] (omniquant.py 432): INFO model.layers.5.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:49 root] (omniquant.py 432): INFO model.layers.5.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:50 root] (omniquant.py 432): INFO model.layers.5.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:51 root] (omniquant.py 432): INFO model.layers.5.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:52 root] (omniquant.py 432): INFO model.layers.5.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:53 root] (omniquant.py 432): INFO model.layers.5.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:54 root] (omniquant.py 432): INFO model.layers.5.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:55 root] (omniquant.py 432): INFO model.layers.5.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:56 root] (omniquant.py 432): INFO model.layers.5.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:10:57 root] (omniquant.py 432): INFO model.layers.5.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 20:27:11 root] (omniquant.py 804): INFO layer 5 iter 0 loss:0.0008505231817252934 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-23 20:42:57 root] (omniquant.py 804): INFO layer 5 iter 1 loss:0.0005954020307399333 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-23 21:03:01 root] (omniquant.py 395): INFO === Start quantize layer 6:6 === +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:17 root] (omniquant.py 432): INFO model.layers.6.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:18 root] (omniquant.py 432): INFO model.layers.6.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:19 root] (omniquant.py 432): INFO model.layers.6.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:20 root] (omniquant.py 432): INFO model.layers.6.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:21 root] (omniquant.py 432): INFO model.layers.6.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:22 root] (omniquant.py 432): INFO model.layers.6.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:23 root] (omniquant.py 432): INFO model.layers.6.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:24 root] (omniquant.py 432): INFO model.layers.6.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:25 root] (omniquant.py 432): INFO model.layers.6.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:26 root] (omniquant.py 432): INFO model.layers.6.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:27 root] (omniquant.py 432): INFO model.layers.6.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:28 root] (omniquant.py 432): INFO model.layers.6.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:28 root] (omniquant.py 432): INFO model.layers.6.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:28 root] (omniquant.py 432): INFO model.layers.6.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:28 root] (omniquant.py 432): INFO model.layers.6.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:28 root] (omniquant.py 432): INFO model.layers.6.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:28 root] (omniquant.py 432): INFO model.layers.6.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:28 root] (omniquant.py 432): INFO model.layers.6.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:28 root] (omniquant.py 432): INFO model.layers.6.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:28 root] (omniquant.py 432): INFO model.layers.6.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:05:28 root] (omniquant.py 432): INFO model.layers.6.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:21:57 root] (omniquant.py 804): INFO layer 6 iter 0 loss:0.001067460048943758 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-23 21:37:39 root] (omniquant.py 804): INFO layer 6 iter 1 loss:0.0009100630995817482 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-23 21:55:53 root] (omniquant.py 395): INFO === Start quantize layer 7:7 === +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:48 root] (omniquant.py 432): INFO model.layers.7.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:49 root] (omniquant.py 432): INFO model.layers.7.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:50 root] (omniquant.py 432): INFO model.layers.7.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:51 root] (omniquant.py 432): INFO model.layers.7.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:52 root] (omniquant.py 432): INFO model.layers.7.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:53 root] (omniquant.py 432): INFO model.layers.7.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:54 root] (omniquant.py 432): INFO model.layers.7.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:55 root] (omniquant.py 432): INFO model.layers.7.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:56 root] (omniquant.py 432): INFO model.layers.7.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:57 root] (omniquant.py 432): INFO model.layers.7.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 21:58:58 root] (omniquant.py 432): INFO model.layers.7.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:14:10 root] (omniquant.py 804): INFO layer 7 iter 0 loss:0.001318949623964727 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-23 22:29:01 root] (omniquant.py 804): INFO layer 7 iter 1 loss:0.0011963616125285625 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-23 22:49:37 root] (omniquant.py 395): INFO === Start quantize layer 8:8 === +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:48 root] (omniquant.py 432): INFO model.layers.8.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:49 root] (omniquant.py 432): INFO model.layers.8.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:50 root] (omniquant.py 432): INFO model.layers.8.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:51 root] (omniquant.py 432): INFO model.layers.8.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:52 root] (omniquant.py 432): INFO model.layers.8.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:53 root] (omniquant.py 432): INFO model.layers.8.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:54 root] (omniquant.py 432): INFO model.layers.8.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:55 root] (omniquant.py 432): INFO model.layers.8.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:56 root] (omniquant.py 432): INFO model.layers.8.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:57 root] (omniquant.py 432): INFO model.layers.8.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 22:51:58 root] (omniquant.py 432): INFO model.layers.8.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:08:06 root] (omniquant.py 804): INFO layer 8 iter 0 loss:0.0016021495684981346 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-23 23:23:44 root] (omniquant.py 804): INFO layer 8 iter 1 loss:0.0015069172950461507 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-23 23:42:08 root] (omniquant.py 395): INFO === Start quantize layer 9:9 === +[2025-03-23 23:44:19 root] (omniquant.py 432): INFO model.layers.9.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:19 root] (omniquant.py 432): INFO model.layers.9.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:19 root] (omniquant.py 432): INFO model.layers.9.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:19 root] (omniquant.py 432): INFO model.layers.9.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:19 root] (omniquant.py 432): INFO model.layers.9.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:20 root] (omniquant.py 432): INFO model.layers.9.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:21 root] (omniquant.py 432): INFO model.layers.9.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:22 root] (omniquant.py 432): INFO model.layers.9.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:23 root] (omniquant.py 432): INFO model.layers.9.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:24 root] (omniquant.py 432): INFO model.layers.9.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:25 root] (omniquant.py 432): INFO model.layers.9.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:26 root] (omniquant.py 432): INFO model.layers.9.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:27 root] (omniquant.py 432): INFO model.layers.9.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:28 root] (omniquant.py 432): INFO model.layers.9.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:29 root] (omniquant.py 432): INFO model.layers.9.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-23 23:44:30 root] (omniquant.py 432): INFO model.layers.9.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:00:40 root] (omniquant.py 804): INFO layer 9 iter 0 loss:0.00243201432749629 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 00:16:32 root] (omniquant.py 804): INFO layer 9 iter 1 loss:0.002052504802122712 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 00:37:12 root] (omniquant.py 395): INFO === Start quantize layer 10:10 === +[2025-03-24 00:39:48 root] (omniquant.py 432): INFO model.layers.10.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:48 root] (omniquant.py 432): INFO model.layers.10.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:48 root] (omniquant.py 432): INFO model.layers.10.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:48 root] (omniquant.py 432): INFO model.layers.10.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:48 root] (omniquant.py 432): INFO model.layers.10.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:48 root] (omniquant.py 432): INFO model.layers.10.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:48 root] (omniquant.py 432): INFO model.layers.10.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:48 root] (omniquant.py 432): INFO model.layers.10.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:49 root] (omniquant.py 432): INFO model.layers.10.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:50 root] (omniquant.py 432): INFO model.layers.10.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:51 root] (omniquant.py 432): INFO model.layers.10.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:52 root] (omniquant.py 432): INFO model.layers.10.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:53 root] (omniquant.py 432): INFO model.layers.10.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:54 root] (omniquant.py 432): INFO model.layers.10.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:55 root] (omniquant.py 432): INFO model.layers.10.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:56 root] (omniquant.py 432): INFO model.layers.10.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:57 root] (omniquant.py 432): INFO model.layers.10.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:58 root] (omniquant.py 432): INFO model.layers.10.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:39:59 root] (omniquant.py 432): INFO model.layers.10.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 00:55:39 root] (omniquant.py 804): INFO layer 10 iter 0 loss:0.0030819487292319536 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 01:10:53 root] (omniquant.py 804): INFO layer 10 iter 1 loss:0.002789749065414071 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 01:32:01 root] (omniquant.py 395): INFO === Start quantize layer 11:11 === +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:15 root] (omniquant.py 432): INFO model.layers.11.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:16 root] (omniquant.py 432): INFO model.layers.11.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:17 root] (omniquant.py 432): INFO model.layers.11.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:18 root] (omniquant.py 432): INFO model.layers.11.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:19 root] (omniquant.py 432): INFO model.layers.11.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:20 root] (omniquant.py 432): INFO model.layers.11.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:21 root] (omniquant.py 432): INFO model.layers.11.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:22 root] (omniquant.py 432): INFO model.layers.11.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:23 root] (omniquant.py 432): INFO model.layers.11.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:24 root] (omniquant.py 432): INFO model.layers.11.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:34:25 root] (omniquant.py 432): INFO model.layers.11.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 01:50:16 root] (omniquant.py 804): INFO layer 11 iter 0 loss:0.0035385307855904102 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 02:05:20 root] (omniquant.py 804): INFO layer 11 iter 1 loss:0.0032423795200884342 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 02:25:46 root] (omniquant.py 395): INFO === Start quantize layer 12:12 === +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:56 root] (omniquant.py 432): INFO model.layers.12.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:57 root] (omniquant.py 432): INFO model.layers.12.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:58 root] (omniquant.py 432): INFO model.layers.12.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:27:59 root] (omniquant.py 432): INFO model.layers.12.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:00 root] (omniquant.py 432): INFO model.layers.12.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:01 root] (omniquant.py 432): INFO model.layers.12.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:02 root] (omniquant.py 432): INFO model.layers.12.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:03 root] (omniquant.py 432): INFO model.layers.12.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:04 root] (omniquant.py 432): INFO model.layers.12.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:05 root] (omniquant.py 432): INFO model.layers.12.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:28:06 root] (omniquant.py 432): INFO model.layers.12.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 02:43:34 root] (omniquant.py 804): INFO layer 12 iter 0 loss:0.003849711501970887 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 02:58:30 root] (omniquant.py 804): INFO layer 12 iter 1 loss:0.0036138654686510563 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 03:20:44 root] (omniquant.py 395): INFO === Start quantize layer 13:13 === +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:53 root] (omniquant.py 432): INFO model.layers.13.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:54 root] (omniquant.py 432): INFO model.layers.13.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:55 root] (omniquant.py 432): INFO model.layers.13.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:56 root] (omniquant.py 432): INFO model.layers.13.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:57 root] (omniquant.py 432): INFO model.layers.13.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:58 root] (omniquant.py 432): INFO model.layers.13.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:22:59 root] (omniquant.py 432): INFO model.layers.13.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:00 root] (omniquant.py 432): INFO model.layers.13.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:01 root] (omniquant.py 432): INFO model.layers.13.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:02 root] (omniquant.py 432): INFO model.layers.13.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:03 root] (omniquant.py 432): INFO model.layers.13.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:23:04 root] (omniquant.py 432): INFO model.layers.13.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 03:39:48 root] (omniquant.py 804): INFO layer 13 iter 0 loss:0.004195965826511383 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 03:56:29 root] (omniquant.py 804): INFO layer 13 iter 1 loss:0.003948185592889786 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 04:18:14 root] (omniquant.py 395): INFO === Start quantize layer 14:14 === +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:24 root] (omniquant.py 432): INFO model.layers.14.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:25 root] (omniquant.py 432): INFO model.layers.14.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:26 root] (omniquant.py 432): INFO model.layers.14.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:27 root] (omniquant.py 432): INFO model.layers.14.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:28 root] (omniquant.py 432): INFO model.layers.14.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:29 root] (omniquant.py 432): INFO model.layers.14.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:30 root] (omniquant.py 432): INFO model.layers.14.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:31 root] (omniquant.py 432): INFO model.layers.14.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:32 root] (omniquant.py 432): INFO model.layers.14.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:33 root] (omniquant.py 432): INFO model.layers.14.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:34 root] (omniquant.py 432): INFO model.layers.14.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:35 root] (omniquant.py 432): INFO model.layers.14.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:35 root] (omniquant.py 432): INFO model.layers.14.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:35 root] (omniquant.py 432): INFO model.layers.14.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:35 root] (omniquant.py 432): INFO model.layers.14.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:35 root] (omniquant.py 432): INFO model.layers.14.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:35 root] (omniquant.py 432): INFO model.layers.14.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:35 root] (omniquant.py 432): INFO model.layers.14.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:35 root] (omniquant.py 432): INFO model.layers.14.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:35 root] (omniquant.py 432): INFO model.layers.14.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:35 root] (omniquant.py 432): INFO model.layers.14.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:20:35 root] (omniquant.py 432): INFO model.layers.14.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 04:37:36 root] (omniquant.py 804): INFO layer 14 iter 0 loss:0.004514928907155991 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 04:54:29 root] (omniquant.py 804): INFO layer 14 iter 1 loss:0.00429201778024435 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 05:15:30 root] (omniquant.py 395): INFO === Start quantize layer 15:15 === +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:41 root] (omniquant.py 432): INFO model.layers.15.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:42 root] (omniquant.py 432): INFO model.layers.15.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:43 root] (omniquant.py 432): INFO model.layers.15.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:44 root] (omniquant.py 432): INFO model.layers.15.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:45 root] (omniquant.py 432): INFO model.layers.15.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:46 root] (omniquant.py 432): INFO model.layers.15.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:47 root] (omniquant.py 432): INFO model.layers.15.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:48 root] (omniquant.py 432): INFO model.layers.15.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:49 root] (omniquant.py 432): INFO model.layers.15.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:50 root] (omniquant.py 432): INFO model.layers.15.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:17:51 root] (omniquant.py 432): INFO model.layers.15.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 05:35:54 root] (omniquant.py 804): INFO layer 15 iter 0 loss:0.005206764210015535 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 05:53:35 root] (omniquant.py 804): INFO layer 15 iter 1 loss:0.004876821767538786 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 06:16:41 root] (omniquant.py 395): INFO === Start quantize layer 16:16 === +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:50 root] (omniquant.py 432): INFO model.layers.16.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:51 root] (omniquant.py 432): INFO model.layers.16.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:52 root] (omniquant.py 432): INFO model.layers.16.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:53 root] (omniquant.py 432): INFO model.layers.16.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:54 root] (omniquant.py 432): INFO model.layers.16.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:55 root] (omniquant.py 432): INFO model.layers.16.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:56 root] (omniquant.py 432): INFO model.layers.16.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:57 root] (omniquant.py 432): INFO model.layers.16.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:58 root] (omniquant.py 432): INFO model.layers.16.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:18:59 root] (omniquant.py 432): INFO model.layers.16.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:19:00 root] (omniquant.py 432): INFO model.layers.16.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 06:35:58 root] (omniquant.py 804): INFO layer 16 iter 0 loss:0.005621984601020813 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 06:52:38 root] (omniquant.py 804): INFO layer 16 iter 1 loss:0.005258884746581316 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 07:15:08 root] (omniquant.py 395): INFO === Start quantize layer 17:17 === +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:18 root] (omniquant.py 432): INFO model.layers.17.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:19 root] (omniquant.py 432): INFO model.layers.17.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:20 root] (omniquant.py 432): INFO model.layers.17.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:21 root] (omniquant.py 432): INFO model.layers.17.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:22 root] (omniquant.py 432): INFO model.layers.17.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:23 root] (omniquant.py 432): INFO model.layers.17.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:24 root] (omniquant.py 432): INFO model.layers.17.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:25 root] (omniquant.py 432): INFO model.layers.17.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:26 root] (omniquant.py 432): INFO model.layers.17.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:27 root] (omniquant.py 432): INFO model.layers.17.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:17:28 root] (omniquant.py 432): INFO model.layers.17.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 07:34:10 root] (omniquant.py 804): INFO layer 17 iter 0 loss:0.006054044701159 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 07:50:07 root] (omniquant.py 804): INFO layer 17 iter 1 loss:0.005778464954346418 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 08:14:02 root] (omniquant.py 395): INFO === Start quantize layer 18:18 === +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:13 root] (omniquant.py 432): INFO model.layers.18.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:14 root] (omniquant.py 432): INFO model.layers.18.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:15 root] (omniquant.py 432): INFO model.layers.18.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:16 root] (omniquant.py 432): INFO model.layers.18.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:17 root] (omniquant.py 432): INFO model.layers.18.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:18 root] (omniquant.py 432): INFO model.layers.18.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:19 root] (omniquant.py 432): INFO model.layers.18.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:20 root] (omniquant.py 432): INFO model.layers.18.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:21 root] (omniquant.py 432): INFO model.layers.18.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:22 root] (omniquant.py 432): INFO model.layers.18.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:16:23 root] (omniquant.py 432): INFO model.layers.18.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 08:33:29 root] (omniquant.py 804): INFO layer 18 iter 0 loss:0.006599904038012028 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 08:50:01 root] (omniquant.py 804): INFO layer 18 iter 1 loss:0.006196401081979275 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 09:13:35 root] (omniquant.py 395): INFO === Start quantize layer 19:19 === +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:43 root] (omniquant.py 432): INFO model.layers.19.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:44 root] (omniquant.py 432): INFO model.layers.19.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:45 root] (omniquant.py 432): INFO model.layers.19.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:46 root] (omniquant.py 432): INFO model.layers.19.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:47 root] (omniquant.py 432): INFO model.layers.19.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:48 root] (omniquant.py 432): INFO model.layers.19.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:49 root] (omniquant.py 432): INFO model.layers.19.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:50 root] (omniquant.py 432): INFO model.layers.19.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:51 root] (omniquant.py 432): INFO model.layers.19.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:52 root] (omniquant.py 432): INFO model.layers.19.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:53 root] (omniquant.py 432): INFO model.layers.19.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:15:54 root] (omniquant.py 432): INFO model.layers.19.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 09:32:40 root] (omniquant.py 804): INFO layer 19 iter 0 loss:0.007031500805169344 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 09:48:58 root] (omniquant.py 804): INFO layer 19 iter 1 loss:0.006604496389627457 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 10:13:07 root] (omniquant.py 395): INFO === Start quantize layer 20:20 === +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:41 root] (omniquant.py 432): INFO model.layers.20.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:42 root] (omniquant.py 432): INFO model.layers.20.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:43 root] (omniquant.py 432): INFO model.layers.20.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:44 root] (omniquant.py 432): INFO model.layers.20.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:45 root] (omniquant.py 432): INFO model.layers.20.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:46 root] (omniquant.py 432): INFO model.layers.20.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:47 root] (omniquant.py 432): INFO model.layers.20.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:48 root] (omniquant.py 432): INFO model.layers.20.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:49 root] (omniquant.py 432): INFO model.layers.20.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:50 root] (omniquant.py 432): INFO model.layers.20.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:51 root] (omniquant.py 432): INFO model.layers.20.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:15:52 root] (omniquant.py 432): INFO model.layers.20.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 10:34:58 root] (omniquant.py 804): INFO layer 20 iter 0 loss:0.007850714027881622 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 10:53:22 root] (omniquant.py 804): INFO layer 20 iter 1 loss:0.00741202337667346 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 11:17:49 root] (omniquant.py 395): INFO === Start quantize layer 21:21 === +[2025-03-24 11:19:55 root] (omniquant.py 432): INFO model.layers.21.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:56 root] (omniquant.py 432): INFO model.layers.21.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:57 root] (omniquant.py 432): INFO model.layers.21.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:58 root] (omniquant.py 432): INFO model.layers.21.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:19:59 root] (omniquant.py 432): INFO model.layers.21.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:00 root] (omniquant.py 432): INFO model.layers.21.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:01 root] (omniquant.py 432): INFO model.layers.21.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:02 root] (omniquant.py 432): INFO model.layers.21.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:03 root] (omniquant.py 432): INFO model.layers.21.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:04 root] (omniquant.py 432): INFO model.layers.21.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:05 root] (omniquant.py 432): INFO model.layers.21.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:20:06 root] (omniquant.py 432): INFO model.layers.21.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 11:37:05 root] (omniquant.py 804): INFO layer 21 iter 0 loss:0.008476762101054192 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 11:52:58 root] (omniquant.py 804): INFO layer 21 iter 1 loss:0.008017047308385372 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 12:16:26 root] (omniquant.py 395): INFO === Start quantize layer 22:22 === +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:31 root] (omniquant.py 432): INFO model.layers.22.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:32 root] (omniquant.py 432): INFO model.layers.22.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:33 root] (omniquant.py 432): INFO model.layers.22.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:34 root] (omniquant.py 432): INFO model.layers.22.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:35 root] (omniquant.py 432): INFO model.layers.22.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:36 root] (omniquant.py 432): INFO model.layers.22.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:37 root] (omniquant.py 432): INFO model.layers.22.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:38 root] (omniquant.py 432): INFO model.layers.22.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:39 root] (omniquant.py 432): INFO model.layers.22.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:40 root] (omniquant.py 432): INFO model.layers.22.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:41 root] (omniquant.py 432): INFO model.layers.22.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:18:42 root] (omniquant.py 432): INFO model.layers.22.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 12:36:09 root] (omniquant.py 804): INFO layer 22 iter 0 loss:0.009470224380493164 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 12:53:29 root] (omniquant.py 804): INFO layer 22 iter 1 loss:0.008972473442554474 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 13:15:14 root] (omniquant.py 395): INFO === Start quantize layer 23:23 === +[2025-03-24 13:17:18 root] (omniquant.py 432): INFO model.layers.23.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:18 root] (omniquant.py 432): INFO model.layers.23.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:18 root] (omniquant.py 432): INFO model.layers.23.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:18 root] (omniquant.py 432): INFO model.layers.23.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:18 root] (omniquant.py 432): INFO model.layers.23.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:19 root] (omniquant.py 432): INFO model.layers.23.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:20 root] (omniquant.py 432): INFO model.layers.23.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:21 root] (omniquant.py 432): INFO model.layers.23.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:22 root] (omniquant.py 432): INFO model.layers.23.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:23 root] (omniquant.py 432): INFO model.layers.23.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:24 root] (omniquant.py 432): INFO model.layers.23.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:25 root] (omniquant.py 432): INFO model.layers.23.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:26 root] (omniquant.py 432): INFO model.layers.23.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:27 root] (omniquant.py 432): INFO model.layers.23.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:28 root] (omniquant.py 432): INFO model.layers.23.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:17:29 root] (omniquant.py 432): INFO model.layers.23.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 13:34:27 root] (omniquant.py 804): INFO layer 23 iter 0 loss:0.010935205966234207 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 13:50:50 root] (omniquant.py 804): INFO layer 23 iter 1 loss:0.010349467396736145 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 14:13:54 root] (omniquant.py 395): INFO === Start quantize layer 24:24 === +[2025-03-24 14:16:08 root] (omniquant.py 432): INFO model.layers.24.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:08 root] (omniquant.py 432): INFO model.layers.24.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:08 root] (omniquant.py 432): INFO model.layers.24.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:08 root] (omniquant.py 432): INFO model.layers.24.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:08 root] (omniquant.py 432): INFO model.layers.24.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:08 root] (omniquant.py 432): INFO model.layers.24.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:08 root] (omniquant.py 432): INFO model.layers.24.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:08 root] (omniquant.py 432): INFO model.layers.24.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:08 root] (omniquant.py 432): INFO model.layers.24.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:09 root] (omniquant.py 432): INFO model.layers.24.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:10 root] (omniquant.py 432): INFO model.layers.24.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:11 root] (omniquant.py 432): INFO model.layers.24.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:12 root] (omniquant.py 432): INFO model.layers.24.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:13 root] (omniquant.py 432): INFO model.layers.24.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:14 root] (omniquant.py 432): INFO model.layers.24.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:15 root] (omniquant.py 432): INFO model.layers.24.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:16 root] (omniquant.py 432): INFO model.layers.24.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:17 root] (omniquant.py 432): INFO model.layers.24.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:18 root] (omniquant.py 432): INFO model.layers.24.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:16:19 root] (omniquant.py 432): INFO model.layers.24.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 14:34:10 root] (omniquant.py 804): INFO layer 24 iter 0 loss:0.012648837640881538 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 14:52:07 root] (omniquant.py 804): INFO layer 24 iter 1 loss:0.012104946188628674 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 15:12:43 root] (omniquant.py 395): INFO === Start quantize layer 25:25 === +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:49 root] (omniquant.py 432): INFO model.layers.25.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:50 root] (omniquant.py 432): INFO model.layers.25.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:51 root] (omniquant.py 432): INFO model.layers.25.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:52 root] (omniquant.py 432): INFO model.layers.25.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:53 root] (omniquant.py 432): INFO model.layers.25.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:54 root] (omniquant.py 432): INFO model.layers.25.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:55 root] (omniquant.py 432): INFO model.layers.25.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:56 root] (omniquant.py 432): INFO model.layers.25.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:57 root] (omniquant.py 432): INFO model.layers.25.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:58 root] (omniquant.py 432): INFO model.layers.25.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:14:59 root] (omniquant.py 432): INFO model.layers.25.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 15:32:22 root] (omniquant.py 804): INFO layer 25 iter 0 loss:0.01396111585199833 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 15:49:33 root] (omniquant.py 804): INFO layer 25 iter 1 loss:0.013415982015430927 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 16:13:38 root] (omniquant.py 395): INFO === Start quantize layer 26:26 === +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:45 root] (omniquant.py 432): INFO model.layers.26.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:46 root] (omniquant.py 432): INFO model.layers.26.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:47 root] (omniquant.py 432): INFO model.layers.26.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:48 root] (omniquant.py 432): INFO model.layers.26.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:49 root] (omniquant.py 432): INFO model.layers.26.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:50 root] (omniquant.py 432): INFO model.layers.26.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:51 root] (omniquant.py 432): INFO model.layers.26.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:52 root] (omniquant.py 432): INFO model.layers.26.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:53 root] (omniquant.py 432): INFO model.layers.26.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:54 root] (omniquant.py 432): INFO model.layers.26.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:15:55 root] (omniquant.py 432): INFO model.layers.26.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 16:32:58 root] (omniquant.py 804): INFO layer 26 iter 0 loss:0.01554612722247839 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 16:49:13 root] (omniquant.py 804): INFO layer 26 iter 1 loss:0.014968768693506718 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 17:15:26 root] (omniquant.py 395): INFO === Start quantize layer 27:27 === +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:31 root] (omniquant.py 432): INFO model.layers.27.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:32 root] (omniquant.py 432): INFO model.layers.27.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:33 root] (omniquant.py 432): INFO model.layers.27.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:34 root] (omniquant.py 432): INFO model.layers.27.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:35 root] (omniquant.py 432): INFO model.layers.27.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:36 root] (omniquant.py 432): INFO model.layers.27.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:37 root] (omniquant.py 432): INFO model.layers.27.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:38 root] (omniquant.py 432): INFO model.layers.27.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:39 root] (omniquant.py 432): INFO model.layers.27.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:40 root] (omniquant.py 432): INFO model.layers.27.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:41 root] (omniquant.py 432): INFO model.layers.27.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:42 root] (omniquant.py 432): INFO model.layers.27.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:42 root] (omniquant.py 432): INFO model.layers.27.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:42 root] (omniquant.py 432): INFO model.layers.27.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:42 root] (omniquant.py 432): INFO model.layers.27.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:42 root] (omniquant.py 432): INFO model.layers.27.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:42 root] (omniquant.py 432): INFO model.layers.27.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:42 root] (omniquant.py 432): INFO model.layers.27.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:42 root] (omniquant.py 432): INFO model.layers.27.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:42 root] (omniquant.py 432): INFO model.layers.27.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:42 root] (omniquant.py 432): INFO model.layers.27.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:42 root] (omniquant.py 432): INFO model.layers.27.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:17:42 root] (omniquant.py 432): INFO model.layers.27.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 17:35:48 root] (omniquant.py 804): INFO layer 27 iter 0 loss:0.01722174882888794 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 17:52:59 root] (omniquant.py 804): INFO layer 27 iter 1 loss:0.016562452539801598 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 18:18:21 root] (omniquant.py 395): INFO === Start quantize layer 28:28 === +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:29 root] (omniquant.py 432): INFO model.layers.28.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:30 root] (omniquant.py 432): INFO model.layers.28.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:31 root] (omniquant.py 432): INFO model.layers.28.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:32 root] (omniquant.py 432): INFO model.layers.28.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:33 root] (omniquant.py 432): INFO model.layers.28.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:34 root] (omniquant.py 432): INFO model.layers.28.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:35 root] (omniquant.py 432): INFO model.layers.28.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:36 root] (omniquant.py 432): INFO model.layers.28.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:37 root] (omniquant.py 432): INFO model.layers.28.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:38 root] (omniquant.py 432): INFO model.layers.28.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:39 root] (omniquant.py 432): INFO model.layers.28.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:40 root] (omniquant.py 432): INFO model.layers.28.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:40 root] (omniquant.py 432): INFO model.layers.28.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:40 root] (omniquant.py 432): INFO model.layers.28.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:40 root] (omniquant.py 432): INFO model.layers.28.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:40 root] (omniquant.py 432): INFO model.layers.28.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:40 root] (omniquant.py 432): INFO model.layers.28.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:40 root] (omniquant.py 432): INFO model.layers.28.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:40 root] (omniquant.py 432): INFO model.layers.28.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:40 root] (omniquant.py 432): INFO model.layers.28.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:40 root] (omniquant.py 432): INFO model.layers.28.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:40 root] (omniquant.py 432): INFO model.layers.28.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:40 root] (omniquant.py 432): INFO model.layers.28.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:20:40 root] (omniquant.py 432): INFO model.layers.28.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 18:38:11 root] (omniquant.py 804): INFO layer 28 iter 0 loss:0.018513668328523636 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 18:54:40 root] (omniquant.py 804): INFO layer 28 iter 1 loss:0.017861129716038704 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 19:22:25 root] (omniquant.py 395): INFO === Start quantize layer 29:29 === +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:31 root] (omniquant.py 432): INFO model.layers.29.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:32 root] (omniquant.py 432): INFO model.layers.29.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:33 root] (omniquant.py 432): INFO model.layers.29.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:34 root] (omniquant.py 432): INFO model.layers.29.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:35 root] (omniquant.py 432): INFO model.layers.29.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:36 root] (omniquant.py 432): INFO model.layers.29.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:37 root] (omniquant.py 432): INFO model.layers.29.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:38 root] (omniquant.py 432): INFO model.layers.29.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:39 root] (omniquant.py 432): INFO model.layers.29.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:40 root] (omniquant.py 432): INFO model.layers.29.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:24:41 root] (omniquant.py 432): INFO model.layers.29.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 19:42:36 root] (omniquant.py 804): INFO layer 29 iter 0 loss:0.01938488893210888 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 19:59:23 root] (omniquant.py 804): INFO layer 29 iter 1 loss:0.018772296607494354 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 20:25:24 root] (omniquant.py 395): INFO === Start quantize layer 30:30 === +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:31 root] (omniquant.py 432): INFO model.layers.30.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:32 root] (omniquant.py 432): INFO model.layers.30.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:33 root] (omniquant.py 432): INFO model.layers.30.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:34 root] (omniquant.py 432): INFO model.layers.30.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:35 root] (omniquant.py 432): INFO model.layers.30.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:36 root] (omniquant.py 432): INFO model.layers.30.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:37 root] (omniquant.py 432): INFO model.layers.30.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:38 root] (omniquant.py 432): INFO model.layers.30.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:39 root] (omniquant.py 432): INFO model.layers.30.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:40 root] (omniquant.py 432): INFO model.layers.30.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:27:41 root] (omniquant.py 432): INFO model.layers.30.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 20:45:39 root] (omniquant.py 804): INFO layer 30 iter 0 loss:0.019629430025815964 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 21:03:09 root] (omniquant.py 804): INFO layer 30 iter 1 loss:0.019011521711945534 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 21:29:57 root] (omniquant.py 395): INFO === Start quantize layer 31:31 === +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:10 root] (omniquant.py 432): INFO model.layers.31.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:11 root] (omniquant.py 432): INFO model.layers.31.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:12 root] (omniquant.py 432): INFO model.layers.31.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:13 root] (omniquant.py 432): INFO model.layers.31.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:14 root] (omniquant.py 432): INFO model.layers.31.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:15 root] (omniquant.py 432): INFO model.layers.31.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:16 root] (omniquant.py 432): INFO model.layers.31.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:17 root] (omniquant.py 432): INFO model.layers.31.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:18 root] (omniquant.py 432): INFO model.layers.31.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:19 root] (omniquant.py 432): INFO model.layers.31.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:32:20 root] (omniquant.py 432): INFO model.layers.31.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 21:50:30 root] (omniquant.py 804): INFO layer 31 iter 0 loss:0.021212389692664146 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 22:08:32 root] (omniquant.py 804): INFO layer 31 iter 1 loss:0.02053677849471569 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 22:35:56 root] (omniquant.py 395): INFO === Start quantize layer 32:32 === +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:08 root] (omniquant.py 432): INFO model.layers.32.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:09 root] (omniquant.py 432): INFO model.layers.32.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:10 root] (omniquant.py 432): INFO model.layers.32.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:11 root] (omniquant.py 432): INFO model.layers.32.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:12 root] (omniquant.py 432): INFO model.layers.32.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:13 root] (omniquant.py 432): INFO model.layers.32.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:14 root] (omniquant.py 432): INFO model.layers.32.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:15 root] (omniquant.py 432): INFO model.layers.32.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:16 root] (omniquant.py 432): INFO model.layers.32.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:17 root] (omniquant.py 432): INFO model.layers.32.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:18 root] (omniquant.py 432): INFO model.layers.32.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:38:19 root] (omniquant.py 432): INFO model.layers.32.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 22:58:14 root] (omniquant.py 804): INFO layer 32 iter 0 loss:0.02227924019098282 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 23:17:12 root] (omniquant.py 804): INFO layer 32 iter 1 loss:0.021494898945093155 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-24 23:44:32 root] (omniquant.py 395): INFO === Start quantize layer 33:33 === +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:45 root] (omniquant.py 432): INFO model.layers.33.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:46 root] (omniquant.py 432): INFO model.layers.33.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:47 root] (omniquant.py 432): INFO model.layers.33.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:48 root] (omniquant.py 432): INFO model.layers.33.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:49 root] (omniquant.py 432): INFO model.layers.33.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:50 root] (omniquant.py 432): INFO model.layers.33.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:51 root] (omniquant.py 432): INFO model.layers.33.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:52 root] (omniquant.py 432): INFO model.layers.33.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:53 root] (omniquant.py 432): INFO model.layers.33.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:54 root] (omniquant.py 432): INFO model.layers.33.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-24 23:46:55 root] (omniquant.py 432): INFO model.layers.33.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:06:03 root] (omniquant.py 804): INFO layer 33 iter 0 loss:0.023341543972492218 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 00:24:06 root] (omniquant.py 804): INFO layer 33 iter 1 loss:0.022636398673057556 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 00:52:58 root] (omniquant.py 395): INFO === Start quantize layer 34:34 === +[2025-03-25 00:55:09 root] (omniquant.py 432): INFO model.layers.34.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:09 root] (omniquant.py 432): INFO model.layers.34.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:09 root] (omniquant.py 432): INFO model.layers.34.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:09 root] (omniquant.py 432): INFO model.layers.34.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:09 root] (omniquant.py 432): INFO model.layers.34.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:09 root] (omniquant.py 432): INFO model.layers.34.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:09 root] (omniquant.py 432): INFO model.layers.34.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:10 root] (omniquant.py 432): INFO model.layers.34.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:11 root] (omniquant.py 432): INFO model.layers.34.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:12 root] (omniquant.py 432): INFO model.layers.34.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:13 root] (omniquant.py 432): INFO model.layers.34.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:14 root] (omniquant.py 432): INFO model.layers.34.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:15 root] (omniquant.py 432): INFO model.layers.34.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:16 root] (omniquant.py 432): INFO model.layers.34.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:17 root] (omniquant.py 432): INFO model.layers.34.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:18 root] (omniquant.py 432): INFO model.layers.34.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:19 root] (omniquant.py 432): INFO model.layers.34.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 00:55:20 root] (omniquant.py 432): INFO model.layers.34.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 01:13:41 root] (omniquant.py 804): INFO layer 34 iter 0 loss:0.024472717195749283 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 01:31:12 root] (omniquant.py 804): INFO layer 34 iter 1 loss:0.023692147806286812 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 02:00:06 root] (omniquant.py 395): INFO === Start quantize layer 35:35 === +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:21 root] (omniquant.py 432): INFO model.layers.35.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:22 root] (omniquant.py 432): INFO model.layers.35.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:23 root] (omniquant.py 432): INFO model.layers.35.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:24 root] (omniquant.py 432): INFO model.layers.35.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:25 root] (omniquant.py 432): INFO model.layers.35.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:26 root] (omniquant.py 432): INFO model.layers.35.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:27 root] (omniquant.py 432): INFO model.layers.35.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:28 root] (omniquant.py 432): INFO model.layers.35.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:29 root] (omniquant.py 432): INFO model.layers.35.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:30 root] (omniquant.py 432): INFO model.layers.35.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:02:31 root] (omniquant.py 432): INFO model.layers.35.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 02:20:42 root] (omniquant.py 804): INFO layer 35 iter 0 loss:0.0255808774381876 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 02:38:31 root] (omniquant.py 804): INFO layer 35 iter 1 loss:0.024863846600055695 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 03:06:29 root] (omniquant.py 395): INFO === Start quantize layer 36:36 === +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:40 root] (omniquant.py 432): INFO model.layers.36.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:41 root] (omniquant.py 432): INFO model.layers.36.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:42 root] (omniquant.py 432): INFO model.layers.36.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:43 root] (omniquant.py 432): INFO model.layers.36.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:44 root] (omniquant.py 432): INFO model.layers.36.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:45 root] (omniquant.py 432): INFO model.layers.36.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:46 root] (omniquant.py 432): INFO model.layers.36.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:47 root] (omniquant.py 432): INFO model.layers.36.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:48 root] (omniquant.py 432): INFO model.layers.36.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:49 root] (omniquant.py 432): INFO model.layers.36.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:50 root] (omniquant.py 432): INFO model.layers.36.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:08:51 root] (omniquant.py 432): INFO model.layers.36.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 03:28:07 root] (omniquant.py 804): INFO layer 36 iter 0 loss:0.026673035696148872 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 03:45:35 root] (omniquant.py 804): INFO layer 36 iter 1 loss:0.025990011170506477 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 04:11:17 root] (omniquant.py 395): INFO === Start quantize layer 37:37 === +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:30 root] (omniquant.py 432): INFO model.layers.37.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:31 root] (omniquant.py 432): INFO model.layers.37.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:32 root] (omniquant.py 432): INFO model.layers.37.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:33 root] (omniquant.py 432): INFO model.layers.37.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:34 root] (omniquant.py 432): INFO model.layers.37.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:35 root] (omniquant.py 432): INFO model.layers.37.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:36 root] (omniquant.py 432): INFO model.layers.37.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:37 root] (omniquant.py 432): INFO model.layers.37.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:38 root] (omniquant.py 432): INFO model.layers.37.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:39 root] (omniquant.py 432): INFO model.layers.37.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:13:40 root] (omniquant.py 432): INFO model.layers.37.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 04:34:09 root] (omniquant.py 804): INFO layer 37 iter 0 loss:0.02724103070795536 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 04:52:25 root] (omniquant.py 804): INFO layer 37 iter 1 loss:0.026567289605736732 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 05:20:22 root] (omniquant.py 395): INFO === Start quantize layer 38:38 === +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:35 root] (omniquant.py 432): INFO model.layers.38.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:36 root] (omniquant.py 432): INFO model.layers.38.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:37 root] (omniquant.py 432): INFO model.layers.38.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:38 root] (omniquant.py 432): INFO model.layers.38.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:39 root] (omniquant.py 432): INFO model.layers.38.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:40 root] (omniquant.py 432): INFO model.layers.38.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:41 root] (omniquant.py 432): INFO model.layers.38.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:42 root] (omniquant.py 432): INFO model.layers.38.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:43 root] (omniquant.py 432): INFO model.layers.38.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:44 root] (omniquant.py 432): INFO model.layers.38.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:45 root] (omniquant.py 432): INFO model.layers.38.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:22:46 root] (omniquant.py 432): INFO model.layers.38.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 05:43:27 root] (omniquant.py 804): INFO layer 38 iter 0 loss:0.028319435194134712 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 06:02:14 root] (omniquant.py 804): INFO layer 38 iter 1 loss:0.027607468888163567 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 06:30:30 root] (omniquant.py 395): INFO === Start quantize layer 39:39 === +[2025-03-25 06:32:38 root] (omniquant.py 432): INFO model.layers.39.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:38 root] (omniquant.py 432): INFO model.layers.39.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:38 root] (omniquant.py 432): INFO model.layers.39.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:38 root] (omniquant.py 432): INFO model.layers.39.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:38 root] (omniquant.py 432): INFO model.layers.39.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:38 root] (omniquant.py 432): INFO model.layers.39.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:38 root] (omniquant.py 432): INFO model.layers.39.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:39 root] (omniquant.py 432): INFO model.layers.39.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:40 root] (omniquant.py 432): INFO model.layers.39.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:41 root] (omniquant.py 432): INFO model.layers.39.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:42 root] (omniquant.py 432): INFO model.layers.39.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:43 root] (omniquant.py 432): INFO model.layers.39.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:44 root] (omniquant.py 432): INFO model.layers.39.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:45 root] (omniquant.py 432): INFO model.layers.39.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:46 root] (omniquant.py 432): INFO model.layers.39.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:47 root] (omniquant.py 432): INFO model.layers.39.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:48 root] (omniquant.py 432): INFO model.layers.39.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:32:49 root] (omniquant.py 432): INFO model.layers.39.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 06:52:10 root] (omniquant.py 804): INFO layer 39 iter 0 loss:0.030245397239923477 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 07:12:09 root] (omniquant.py 804): INFO layer 39 iter 1 loss:0.02943499945104122 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 07:41:39 root] (omniquant.py 395): INFO === Start quantize layer 40:40 === +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:53 root] (omniquant.py 432): INFO model.layers.40.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:54 root] (omniquant.py 432): INFO model.layers.40.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:55 root] (omniquant.py 432): INFO model.layers.40.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:56 root] (omniquant.py 432): INFO model.layers.40.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:57 root] (omniquant.py 432): INFO model.layers.40.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:58 root] (omniquant.py 432): INFO model.layers.40.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:43:59 root] (omniquant.py 432): INFO model.layers.40.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:00 root] (omniquant.py 432): INFO model.layers.40.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:01 root] (omniquant.py 432): INFO model.layers.40.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:02 root] (omniquant.py 432): INFO model.layers.40.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 07:44:03 root] (omniquant.py 432): INFO model.layers.40.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:03:29 root] (omniquant.py 804): INFO layer 40 iter 0 loss:0.03197420760989189 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 08:22:07 root] (omniquant.py 804): INFO layer 40 iter 1 loss:0.031096015125513077 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 08:52:50 root] (omniquant.py 395): INFO === Start quantize layer 41:41 === +[2025-03-25 08:55:04 root] (omniquant.py 432): INFO model.layers.41.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:04 root] (omniquant.py 432): INFO model.layers.41.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:04 root] (omniquant.py 432): INFO model.layers.41.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:05 root] (omniquant.py 432): INFO model.layers.41.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:06 root] (omniquant.py 432): INFO model.layers.41.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:07 root] (omniquant.py 432): INFO model.layers.41.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:08 root] (omniquant.py 432): INFO model.layers.41.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:09 root] (omniquant.py 432): INFO model.layers.41.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:10 root] (omniquant.py 432): INFO model.layers.41.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:11 root] (omniquant.py 432): INFO model.layers.41.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:12 root] (omniquant.py 432): INFO model.layers.41.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:13 root] (omniquant.py 432): INFO model.layers.41.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:14 root] (omniquant.py 432): INFO model.layers.41.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 08:55:15 root] (omniquant.py 432): INFO model.layers.41.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 09:14:12 root] (omniquant.py 804): INFO layer 41 iter 0 loss:0.033071547746658325 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 09:32:57 root] (omniquant.py 804): INFO layer 41 iter 1 loss:0.03223971650004387 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 10:03:21 root] (omniquant.py 395): INFO === Start quantize layer 42:42 === +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:34 root] (omniquant.py 432): INFO model.layers.42.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:35 root] (omniquant.py 432): INFO model.layers.42.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:36 root] (omniquant.py 432): INFO model.layers.42.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:37 root] (omniquant.py 432): INFO model.layers.42.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:38 root] (omniquant.py 432): INFO model.layers.42.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:39 root] (omniquant.py 432): INFO model.layers.42.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:40 root] (omniquant.py 432): INFO model.layers.42.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:41 root] (omniquant.py 432): INFO model.layers.42.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:42 root] (omniquant.py 432): INFO model.layers.42.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:43 root] (omniquant.py 432): INFO model.layers.42.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:44 root] (omniquant.py 432): INFO model.layers.42.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:45 root] (omniquant.py 432): INFO model.layers.42.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:05:45 root] (omniquant.py 432): INFO model.layers.42.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 10:24:47 root] (omniquant.py 804): INFO layer 42 iter 0 loss:0.0347314290702343 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 10:42:02 root] (omniquant.py 804): INFO layer 42 iter 1 loss:0.033870577812194824 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 11:11:15 root] (omniquant.py 395): INFO === Start quantize layer 43:43 === +[2025-03-25 11:13:28 root] (omniquant.py 432): INFO model.layers.43.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:28 root] (omniquant.py 432): INFO model.layers.43.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:28 root] (omniquant.py 432): INFO model.layers.43.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:28 root] (omniquant.py 432): INFO model.layers.43.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:28 root] (omniquant.py 432): INFO model.layers.43.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:28 root] (omniquant.py 432): INFO model.layers.43.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:28 root] (omniquant.py 432): INFO model.layers.43.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:28 root] (omniquant.py 432): INFO model.layers.43.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:28 root] (omniquant.py 432): INFO model.layers.43.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:28 root] (omniquant.py 432): INFO model.layers.43.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:28 root] (omniquant.py 432): INFO model.layers.43.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:28 root] (omniquant.py 432): INFO model.layers.43.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:29 root] (omniquant.py 432): INFO model.layers.43.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:30 root] (omniquant.py 432): INFO model.layers.43.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:31 root] (omniquant.py 432): INFO model.layers.43.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:32 root] (omniquant.py 432): INFO model.layers.43.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:33 root] (omniquant.py 432): INFO model.layers.43.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:34 root] (omniquant.py 432): INFO model.layers.43.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:35 root] (omniquant.py 432): INFO model.layers.43.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:36 root] (omniquant.py 432): INFO model.layers.43.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:37 root] (omniquant.py 432): INFO model.layers.43.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:38 root] (omniquant.py 432): INFO model.layers.43.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:13:39 root] (omniquant.py 432): INFO model.layers.43.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 11:32:31 root] (omniquant.py 804): INFO layer 43 iter 0 loss:0.03562535345554352 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 11:50:19 root] (omniquant.py 804): INFO layer 43 iter 1 loss:0.03489484265446663 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 12:21:53 root] (omniquant.py 395): INFO === Start quantize layer 44:44 === +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:01 root] (omniquant.py 432): INFO model.layers.44.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:02 root] (omniquant.py 432): INFO model.layers.44.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:03 root] (omniquant.py 432): INFO model.layers.44.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:04 root] (omniquant.py 432): INFO model.layers.44.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:05 root] (omniquant.py 432): INFO model.layers.44.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:06 root] (omniquant.py 432): INFO model.layers.44.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:07 root] (omniquant.py 432): INFO model.layers.44.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:08 root] (omniquant.py 432): INFO model.layers.44.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:09 root] (omniquant.py 432): INFO model.layers.44.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:10 root] (omniquant.py 432): INFO model.layers.44.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:11 root] (omniquant.py 432): INFO model.layers.44.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:24:12 root] (omniquant.py 432): INFO model.layers.44.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 12:43:48 root] (omniquant.py 804): INFO layer 44 iter 0 loss:0.036682166159152985 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 13:03:05 root] (omniquant.py 804): INFO layer 44 iter 1 loss:0.03560290113091469 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 13:34:03 root] (omniquant.py 395): INFO === Start quantize layer 45:45 === +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:13 root] (omniquant.py 432): INFO model.layers.45.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:14 root] (omniquant.py 432): INFO model.layers.45.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:15 root] (omniquant.py 432): INFO model.layers.45.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:16 root] (omniquant.py 432): INFO model.layers.45.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:17 root] (omniquant.py 432): INFO model.layers.45.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:18 root] (omniquant.py 432): INFO model.layers.45.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:19 root] (omniquant.py 432): INFO model.layers.45.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:20 root] (omniquant.py 432): INFO model.layers.45.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:21 root] (omniquant.py 432): INFO model.layers.45.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:22 root] (omniquant.py 432): INFO model.layers.45.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:36:23 root] (omniquant.py 432): INFO model.layers.45.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 13:56:35 root] (omniquant.py 804): INFO layer 45 iter 0 loss:0.037818875163793564 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 14:14:42 root] (omniquant.py 804): INFO layer 45 iter 1 loss:0.036769624799489975 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 14:45:00 root] (omniquant.py 395): INFO === Start quantize layer 46:46 === +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:12 root] (omniquant.py 432): INFO model.layers.46.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:13 root] (omniquant.py 432): INFO model.layers.46.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:14 root] (omniquant.py 432): INFO model.layers.46.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:15 root] (omniquant.py 432): INFO model.layers.46.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:16 root] (omniquant.py 432): INFO model.layers.46.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:17 root] (omniquant.py 432): INFO model.layers.46.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:18 root] (omniquant.py 432): INFO model.layers.46.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:19 root] (omniquant.py 432): INFO model.layers.46.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:20 root] (omniquant.py 432): INFO model.layers.46.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:21 root] (omniquant.py 432): INFO model.layers.46.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:22 root] (omniquant.py 432): INFO model.layers.46.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 14:47:23 root] (omniquant.py 432): INFO model.layers.46.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:06:00 root] (omniquant.py 804): INFO layer 46 iter 0 loss:0.038852572441101074 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 15:23:44 root] (omniquant.py 804): INFO layer 46 iter 1 loss:0.038002971559762955 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 15:54:16 root] (omniquant.py 395): INFO === Start quantize layer 47:47 === +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:27 root] (omniquant.py 432): INFO model.layers.47.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:28 root] (omniquant.py 432): INFO model.layers.47.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:29 root] (omniquant.py 432): INFO model.layers.47.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:30 root] (omniquant.py 432): INFO model.layers.47.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:31 root] (omniquant.py 432): INFO model.layers.47.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:32 root] (omniquant.py 432): INFO model.layers.47.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:33 root] (omniquant.py 432): INFO model.layers.47.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:34 root] (omniquant.py 432): INFO model.layers.47.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:35 root] (omniquant.py 432): INFO model.layers.47.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:36 root] (omniquant.py 432): INFO model.layers.47.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:37 root] (omniquant.py 432): INFO model.layers.47.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:38 root] (omniquant.py 432): INFO model.layers.47.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:38 root] (omniquant.py 432): INFO model.layers.47.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:38 root] (omniquant.py 432): INFO model.layers.47.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:38 root] (omniquant.py 432): INFO model.layers.47.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:38 root] (omniquant.py 432): INFO model.layers.47.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:38 root] (omniquant.py 432): INFO model.layers.47.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:38 root] (omniquant.py 432): INFO model.layers.47.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:38 root] (omniquant.py 432): INFO model.layers.47.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:38 root] (omniquant.py 432): INFO model.layers.47.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:38 root] (omniquant.py 432): INFO model.layers.47.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:38 root] (omniquant.py 432): INFO model.layers.47.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 15:56:38 root] (omniquant.py 432): INFO model.layers.47.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 16:15:43 root] (omniquant.py 804): INFO layer 47 iter 0 loss:0.040445275604724884 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 16:34:12 root] (omniquant.py 804): INFO layer 47 iter 1 loss:0.03948120027780533 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 17:03:42 root] (omniquant.py 395): INFO === Start quantize layer 48:48 === +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:02 root] (omniquant.py 432): INFO model.layers.48.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:03 root] (omniquant.py 432): INFO model.layers.48.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:04 root] (omniquant.py 432): INFO model.layers.48.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:05 root] (omniquant.py 432): INFO model.layers.48.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:06 root] (omniquant.py 432): INFO model.layers.48.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:07 root] (omniquant.py 432): INFO model.layers.48.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:08 root] (omniquant.py 432): INFO model.layers.48.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:09 root] (omniquant.py 432): INFO model.layers.48.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:10 root] (omniquant.py 432): INFO model.layers.48.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:11 root] (omniquant.py 432): INFO model.layers.48.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:12 root] (omniquant.py 432): INFO model.layers.48.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:06:13 root] (omniquant.py 432): INFO model.layers.48.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 17:25:10 root] (omniquant.py 804): INFO layer 48 iter 0 loss:0.04188726097345352 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 17:43:13 root] (omniquant.py 804): INFO layer 48 iter 1 loss:0.04080546274781227 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 18:13:46 root] (omniquant.py 395): INFO === Start quantize layer 49:49 === +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:56 root] (omniquant.py 432): INFO model.layers.49.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:57 root] (omniquant.py 432): INFO model.layers.49.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:58 root] (omniquant.py 432): INFO model.layers.49.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:15:59 root] (omniquant.py 432): INFO model.layers.49.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:00 root] (omniquant.py 432): INFO model.layers.49.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:01 root] (omniquant.py 432): INFO model.layers.49.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:02 root] (omniquant.py 432): INFO model.layers.49.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:03 root] (omniquant.py 432): INFO model.layers.49.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:04 root] (omniquant.py 432): INFO model.layers.49.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:05 root] (omniquant.py 432): INFO model.layers.49.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:06 root] (omniquant.py 432): INFO model.layers.49.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:16:07 root] (omniquant.py 432): INFO model.layers.49.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 18:36:47 root] (omniquant.py 804): INFO layer 49 iter 0 loss:0.042898230254650116 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 18:56:42 root] (omniquant.py 804): INFO layer 49 iter 1 loss:0.04164396971464157 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 19:27:51 root] (omniquant.py 395): INFO === Start quantize layer 50:50 === +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:04 root] (omniquant.py 432): INFO model.layers.50.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:05 root] (omniquant.py 432): INFO model.layers.50.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:06 root] (omniquant.py 432): INFO model.layers.50.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:07 root] (omniquant.py 432): INFO model.layers.50.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:08 root] (omniquant.py 432): INFO model.layers.50.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:09 root] (omniquant.py 432): INFO model.layers.50.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:10 root] (omniquant.py 432): INFO model.layers.50.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:11 root] (omniquant.py 432): INFO model.layers.50.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:12 root] (omniquant.py 432): INFO model.layers.50.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:13 root] (omniquant.py 432): INFO model.layers.50.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:30:14 root] (omniquant.py 432): INFO model.layers.50.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 19:50:02 root] (omniquant.py 804): INFO layer 50 iter 0 loss:0.04349417984485626 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 20:08:43 root] (omniquant.py 804): INFO layer 50 iter 1 loss:0.042544424533843994 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 20:39:50 root] (omniquant.py 395): INFO === Start quantize layer 51:51 === +[2025-03-25 20:42:04 root] (omniquant.py 432): INFO model.layers.51.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:04 root] (omniquant.py 432): INFO model.layers.51.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:04 root] (omniquant.py 432): INFO model.layers.51.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:04 root] (omniquant.py 432): INFO model.layers.51.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:04 root] (omniquant.py 432): INFO model.layers.51.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:04 root] (omniquant.py 432): INFO model.layers.51.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:04 root] (omniquant.py 432): INFO model.layers.51.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:04 root] (omniquant.py 432): INFO model.layers.51.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:04 root] (omniquant.py 432): INFO model.layers.51.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:04 root] (omniquant.py 432): INFO model.layers.51.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:04 root] (omniquant.py 432): INFO model.layers.51.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:05 root] (omniquant.py 432): INFO model.layers.51.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:06 root] (omniquant.py 432): INFO model.layers.51.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:07 root] (omniquant.py 432): INFO model.layers.51.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:08 root] (omniquant.py 432): INFO model.layers.51.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:09 root] (omniquant.py 432): INFO model.layers.51.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:10 root] (omniquant.py 432): INFO model.layers.51.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:11 root] (omniquant.py 432): INFO model.layers.51.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:12 root] (omniquant.py 432): INFO model.layers.51.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:13 root] (omniquant.py 432): INFO model.layers.51.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:14 root] (omniquant.py 432): INFO model.layers.51.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 20:42:15 root] (omniquant.py 432): INFO model.layers.51.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:02:16 root] (omniquant.py 804): INFO layer 51 iter 0 loss:0.04343944042921066 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 21:21:50 root] (omniquant.py 804): INFO layer 51 iter 1 loss:0.04239724949002266 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 21:54:17 root] (omniquant.py 395): INFO === Start quantize layer 52:52 === +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:34 root] (omniquant.py 432): INFO model.layers.52.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:35 root] (omniquant.py 432): INFO model.layers.52.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:36 root] (omniquant.py 432): INFO model.layers.52.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:37 root] (omniquant.py 432): INFO model.layers.52.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:38 root] (omniquant.py 432): INFO model.layers.52.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:39 root] (omniquant.py 432): INFO model.layers.52.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:40 root] (omniquant.py 432): INFO model.layers.52.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:41 root] (omniquant.py 432): INFO model.layers.52.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:42 root] (omniquant.py 432): INFO model.layers.52.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:43 root] (omniquant.py 432): INFO model.layers.52.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 21:56:44 root] (omniquant.py 432): INFO model.layers.52.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 22:16:44 root] (omniquant.py 804): INFO layer 52 iter 0 loss:0.043179478496313095 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 22:35:33 root] (omniquant.py 804): INFO layer 52 iter 1 loss:0.04212117940187454 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 23:08:17 root] (omniquant.py 395): INFO === Start quantize layer 53:53 === +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:31 root] (omniquant.py 432): INFO model.layers.53.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:32 root] (omniquant.py 432): INFO model.layers.53.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:33 root] (omniquant.py 432): INFO model.layers.53.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:34 root] (omniquant.py 432): INFO model.layers.53.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:35 root] (omniquant.py 432): INFO model.layers.53.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:36 root] (omniquant.py 432): INFO model.layers.53.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:37 root] (omniquant.py 432): INFO model.layers.53.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:38 root] (omniquant.py 432): INFO model.layers.53.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:39 root] (omniquant.py 432): INFO model.layers.53.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:40 root] (omniquant.py 432): INFO model.layers.53.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:41 root] (omniquant.py 432): INFO model.layers.53.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:42 root] (omniquant.py 432): INFO model.layers.53.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:42 root] (omniquant.py 432): INFO model.layers.53.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:42 root] (omniquant.py 432): INFO model.layers.53.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:42 root] (omniquant.py 432): INFO model.layers.53.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:42 root] (omniquant.py 432): INFO model.layers.53.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:42 root] (omniquant.py 432): INFO model.layers.53.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:42 root] (omniquant.py 432): INFO model.layers.53.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:42 root] (omniquant.py 432): INFO model.layers.53.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:42 root] (omniquant.py 432): INFO model.layers.53.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:42 root] (omniquant.py 432): INFO model.layers.53.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:42 root] (omniquant.py 432): INFO model.layers.53.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:10:42 root] (omniquant.py 432): INFO model.layers.53.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-25 23:29:36 root] (omniquant.py 804): INFO layer 53 iter 0 loss:0.04246075823903084 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-25 23:47:38 root] (omniquant.py 804): INFO layer 53 iter 1 loss:0.04151882976293564 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 00:20:45 root] (omniquant.py 395): INFO === Start quantize layer 54:54 === +[2025-03-26 00:22:56 root] (omniquant.py 432): INFO model.layers.54.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:56 root] (omniquant.py 432): INFO model.layers.54.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:56 root] (omniquant.py 432): INFO model.layers.54.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:56 root] (omniquant.py 432): INFO model.layers.54.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:57 root] (omniquant.py 432): INFO model.layers.54.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:58 root] (omniquant.py 432): INFO model.layers.54.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:22:59 root] (omniquant.py 432): INFO model.layers.54.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:00 root] (omniquant.py 432): INFO model.layers.54.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:01 root] (omniquant.py 432): INFO model.layers.54.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:02 root] (omniquant.py 432): INFO model.layers.54.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:03 root] (omniquant.py 432): INFO model.layers.54.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:04 root] (omniquant.py 432): INFO model.layers.54.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:05 root] (omniquant.py 432): INFO model.layers.54.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:06 root] (omniquant.py 432): INFO model.layers.54.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:23:07 root] (omniquant.py 432): INFO model.layers.54.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 00:43:14 root] (omniquant.py 804): INFO layer 54 iter 0 loss:0.041908323764801025 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 01:02:13 root] (omniquant.py 804): INFO layer 54 iter 1 loss:0.04099615290760994 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 01:36:18 root] (omniquant.py 395): INFO === Start quantize layer 55:55 === +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:34 root] (omniquant.py 432): INFO model.layers.55.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:35 root] (omniquant.py 432): INFO model.layers.55.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:36 root] (omniquant.py 432): INFO model.layers.55.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:37 root] (omniquant.py 432): INFO model.layers.55.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:38 root] (omniquant.py 432): INFO model.layers.55.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:39 root] (omniquant.py 432): INFO model.layers.55.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:40 root] (omniquant.py 432): INFO model.layers.55.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:41 root] (omniquant.py 432): INFO model.layers.55.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:42 root] (omniquant.py 432): INFO model.layers.55.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:43 root] (omniquant.py 432): INFO model.layers.55.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:44 root] (omniquant.py 432): INFO model.layers.55.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:45 root] (omniquant.py 432): INFO model.layers.55.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:45 root] (omniquant.py 432): INFO model.layers.55.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:45 root] (omniquant.py 432): INFO model.layers.55.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:45 root] (omniquant.py 432): INFO model.layers.55.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:45 root] (omniquant.py 432): INFO model.layers.55.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:45 root] (omniquant.py 432): INFO model.layers.55.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:45 root] (omniquant.py 432): INFO model.layers.55.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:45 root] (omniquant.py 432): INFO model.layers.55.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:45 root] (omniquant.py 432): INFO model.layers.55.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:38:45 root] (omniquant.py 432): INFO model.layers.55.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 01:58:21 root] (omniquant.py 804): INFO layer 55 iter 0 loss:0.04135853424668312 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 02:16:42 root] (omniquant.py 804): INFO layer 55 iter 1 loss:0.040340568870306015 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 02:50:31 root] (omniquant.py 395): INFO === Start quantize layer 56:56 === +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:47 root] (omniquant.py 432): INFO model.layers.56.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:48 root] (omniquant.py 432): INFO model.layers.56.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:49 root] (omniquant.py 432): INFO model.layers.56.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:50 root] (omniquant.py 432): INFO model.layers.56.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:51 root] (omniquant.py 432): INFO model.layers.56.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:52 root] (omniquant.py 432): INFO model.layers.56.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:53 root] (omniquant.py 432): INFO model.layers.56.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:54 root] (omniquant.py 432): INFO model.layers.56.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:55 root] (omniquant.py 432): INFO model.layers.56.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:56 root] (omniquant.py 432): INFO model.layers.56.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 02:52:57 root] (omniquant.py 432): INFO model.layers.56.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 03:12:34 root] (omniquant.py 804): INFO layer 56 iter 0 loss:0.04278020188212395 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 03:30:55 root] (omniquant.py 804): INFO layer 56 iter 1 loss:0.04191109165549278 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 04:02:44 root] (omniquant.py 395): INFO === Start quantize layer 57:57 === +[2025-03-26 04:04:57 root] (omniquant.py 432): INFO model.layers.57.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:57 root] (omniquant.py 432): INFO model.layers.57.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:57 root] (omniquant.py 432): INFO model.layers.57.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:57 root] (omniquant.py 432): INFO model.layers.57.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:57 root] (omniquant.py 432): INFO model.layers.57.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:58 root] (omniquant.py 432): INFO model.layers.57.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:04:59 root] (omniquant.py 432): INFO model.layers.57.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:00 root] (omniquant.py 432): INFO model.layers.57.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:01 root] (omniquant.py 432): INFO model.layers.57.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:02 root] (omniquant.py 432): INFO model.layers.57.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:03 root] (omniquant.py 432): INFO model.layers.57.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:04 root] (omniquant.py 432): INFO model.layers.57.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:05 root] (omniquant.py 432): INFO model.layers.57.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:06 root] (omniquant.py 432): INFO model.layers.57.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:07 root] (omniquant.py 432): INFO model.layers.57.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:05:08 root] (omniquant.py 432): INFO model.layers.57.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 04:25:08 root] (omniquant.py 804): INFO layer 57 iter 0 loss:0.04395303875207901 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 04:44:07 root] (omniquant.py 804): INFO layer 57 iter 1 loss:0.043154746294021606 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 05:18:27 root] (omniquant.py 395): INFO === Start quantize layer 58:58 === +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:39 root] (omniquant.py 432): INFO model.layers.58.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:40 root] (omniquant.py 432): INFO model.layers.58.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:41 root] (omniquant.py 432): INFO model.layers.58.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:42 root] (omniquant.py 432): INFO model.layers.58.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:43 root] (omniquant.py 432): INFO model.layers.58.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:44 root] (omniquant.py 432): INFO model.layers.58.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:45 root] (omniquant.py 432): INFO model.layers.58.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:46 root] (omniquant.py 432): INFO model.layers.58.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:47 root] (omniquant.py 432): INFO model.layers.58.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:48 root] (omniquant.py 432): INFO model.layers.58.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:49 root] (omniquant.py 432): INFO model.layers.58.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:50 root] (omniquant.py 432): INFO model.layers.58.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:50 root] (omniquant.py 432): INFO model.layers.58.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:50 root] (omniquant.py 432): INFO model.layers.58.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:50 root] (omniquant.py 432): INFO model.layers.58.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:50 root] (omniquant.py 432): INFO model.layers.58.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:50 root] (omniquant.py 432): INFO model.layers.58.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:50 root] (omniquant.py 432): INFO model.layers.58.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:50 root] (omniquant.py 432): INFO model.layers.58.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:20:50 root] (omniquant.py 432): INFO model.layers.58.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 05:40:11 root] (omniquant.py 804): INFO layer 58 iter 0 loss:0.04526657983660698 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 05:58:35 root] (omniquant.py 804): INFO layer 58 iter 1 loss:0.04446420073509216 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 06:31:26 root] (omniquant.py 395): INFO === Start quantize layer 59:59 === +[2025-03-26 06:33:38 root] (omniquant.py 432): INFO model.layers.59.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:38 root] (omniquant.py 432): INFO model.layers.59.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:38 root] (omniquant.py 432): INFO model.layers.59.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:38 root] (omniquant.py 432): INFO model.layers.59.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:39 root] (omniquant.py 432): INFO model.layers.59.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:40 root] (omniquant.py 432): INFO model.layers.59.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:41 root] (omniquant.py 432): INFO model.layers.59.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:42 root] (omniquant.py 432): INFO model.layers.59.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:43 root] (omniquant.py 432): INFO model.layers.59.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:44 root] (omniquant.py 432): INFO model.layers.59.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:45 root] (omniquant.py 432): INFO model.layers.59.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:46 root] (omniquant.py 432): INFO model.layers.59.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:47 root] (omniquant.py 432): INFO model.layers.59.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:48 root] (omniquant.py 432): INFO model.layers.59.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:33:49 root] (omniquant.py 432): INFO model.layers.59.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 06:54:03 root] (omniquant.py 804): INFO layer 59 iter 0 loss:0.10598651319742203 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 07:13:04 root] (omniquant.py 804): INFO layer 59 iter 1 loss:0.09793747961521149 fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 07:46:06 root] (omniquant.py 395): INFO === Start quantize layer 60:60 === +[2025-03-26 07:48:23 root] (omniquant.py 432): INFO model.layers.60.self_attn.q_a_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:23 root] (omniquant.py 432): INFO model.layers.60.self_attn.q_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:23 root] (omniquant.py 432): INFO model.layers.60.self_attn.kv_a_proj_with_mqa: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:23 root] (omniquant.py 432): INFO model.layers.60.self_attn.kv_b_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:23 root] (omniquant.py 432): INFO model.layers.60.self_attn.o_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:23 root] (omniquant.py 432): INFO model.layers.60.experts.0.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.0.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.0.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.1.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.1.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.1.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.2.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.2.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.2.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.3.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.3.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.3.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.4.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.4.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.4.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.5.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.5.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.5.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.6.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.6.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.6.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.7.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.7.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.7.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.8.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.8.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.8.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.9.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.9.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.9.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.10.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.10.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.10.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.11.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.11.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.11.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.12.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.12.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.12.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.13.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.13.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.13.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.14.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.14.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.14.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.15.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.15.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.15.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.16.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.16.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.16.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.17.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.17.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.17.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.18.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.18.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.18.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.19.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.19.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.19.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.20.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.20.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.20.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.21.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.21.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.21.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.22.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.22.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.22.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.23.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.23.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.23.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.24.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.24.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:24 root] (omniquant.py 432): INFO model.layers.60.experts.24.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.25.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.25.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.25.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.26.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.26.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.26.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.27.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.27.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.27.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.28.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.28.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.28.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.29.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.29.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.29.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.30.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.30.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.30.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.31.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.31.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.31.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.32.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.32.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.32.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.33.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.33.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.33.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.34.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.34.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.34.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.35.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.35.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.35.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.36.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.36.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.36.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.37.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.37.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.37.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.38.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.38.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.38.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.39.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.39.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.39.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.40.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.40.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.40.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.41.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.41.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.41.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.42.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.42.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.42.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.43.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.43.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.43.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.44.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.44.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.44.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.45.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.45.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.45.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.46.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.46.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.46.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.47.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.47.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.47.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.48.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.48.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.48.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:25 root] (omniquant.py 432): INFO model.layers.60.experts.49.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.49.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.49.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.50.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.50.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.50.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.51.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.51.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.51.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.52.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.52.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.52.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.53.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.53.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.53.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.54.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.54.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.54.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.55.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.55.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.55.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.56.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.56.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.56.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.57.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.57.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.57.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.58.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.58.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.58.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.59.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.59.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.59.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.60.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.60.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.60.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.61.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.61.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.61.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.62.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.62.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.62.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.63.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.63.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.63.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.64.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.64.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.64.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.65.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.65.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.65.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.66.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.66.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.66.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.67.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.67.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.67.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.68.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.68.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.68.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.69.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.69.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.69.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.70.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.70.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.70.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.71.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.71.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.71.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.72.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.72.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.72.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.73.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.73.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:26 root] (omniquant.py 432): INFO model.layers.60.experts.73.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.74.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.74.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.74.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.75.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.75.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.75.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.76.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.76.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.76.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.77.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.77.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.77.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.78.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.78.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.78.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.79.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.79.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.79.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.80.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.80.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.80.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.81.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.81.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.81.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.82.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.82.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.82.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.83.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.83.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.83.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.84.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.84.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.84.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.85.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.85.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.85.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.86.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.86.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.86.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.87.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.87.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.87.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.88.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.88.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.88.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.89.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.89.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.89.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.90.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.90.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.90.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.91.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.91.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.91.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.92.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.92.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.92.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.93.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.93.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.93.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.94.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.94.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.94.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.95.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.95.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.95.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.96.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.96.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.96.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.97.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.97.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.97.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:27 root] (omniquant.py 432): INFO model.layers.60.experts.98.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.98.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.98.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.99.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.99.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.99.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.100.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.100.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.100.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.101.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.101.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.101.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.102.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.102.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.102.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.103.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.103.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.103.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.104.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.104.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.104.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.105.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.105.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.105.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.106.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.106.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.106.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.107.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.107.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.107.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.108.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.108.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.108.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.109.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.109.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.109.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.110.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.110.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.110.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.111.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.111.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.111.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.112.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.112.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.112.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.113.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.113.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.113.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.114.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.114.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.114.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.115.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.115.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.115.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.116.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.116.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.116.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.117.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.117.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.117.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.118.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.118.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.118.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.119.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.119.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.119.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.120.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.120.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.120.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.121.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.121.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.121.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.122.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:28 root] (omniquant.py 432): INFO model.layers.60.experts.122.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.122.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.123.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.123.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.123.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.124.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.124.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.124.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.125.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.125.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.125.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.126.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.126.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.126.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.127.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.127.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.127.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.128.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.128.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.128.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.129.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.129.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.129.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.130.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.130.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.130.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.131.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.131.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.131.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.132.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.132.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.132.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.133.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.133.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.133.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.134.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.134.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.134.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.135.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.135.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.135.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.136.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.136.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.136.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.137.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.137.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.137.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.138.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.138.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.138.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.139.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.139.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.139.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.140.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.140.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.140.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.141.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.141.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.141.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.142.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.142.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.142.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.143.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.143.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.143.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.144.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.144.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.144.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.145.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.145.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.145.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.146.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.146.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:29 root] (omniquant.py 432): INFO model.layers.60.experts.146.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.147.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.147.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.147.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.148.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.148.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.148.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.149.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.149.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.149.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.150.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.150.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.150.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.151.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.151.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.151.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.152.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.152.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.152.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.153.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.153.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.153.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.154.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.154.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.154.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.155.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.155.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.155.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.156.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.156.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.156.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.157.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.157.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.157.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.158.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.158.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.158.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.159.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.159.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.159.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.160.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.160.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.160.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.161.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.161.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.161.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.162.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.162.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.162.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.163.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.163.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.163.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.164.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.164.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.164.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.165.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.165.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.165.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.166.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.166.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.166.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.167.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.167.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.167.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.168.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.168.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.168.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.169.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.169.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.169.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.170.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.170.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.170.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:30 root] (omniquant.py 432): INFO model.layers.60.experts.171.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.171.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.171.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.172.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.172.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.172.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.173.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.173.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.173.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.174.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.174.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.174.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.175.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.175.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.175.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.176.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.176.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.176.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.177.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.177.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.177.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.178.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.178.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.178.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.179.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.179.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.179.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.180.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.180.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.180.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.181.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.181.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.181.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.182.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.182.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.182.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.183.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.183.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.183.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.184.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.184.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.184.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.185.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.185.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.185.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.186.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.186.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.186.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.187.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.187.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.187.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.188.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.188.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.188.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.189.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.189.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.189.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.190.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.190.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.190.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.191.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.191.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.191.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.192.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.192.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.192.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.193.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.193.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.193.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.194.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.194.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.194.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.195.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:31 root] (omniquant.py 432): INFO model.layers.60.experts.195.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.195.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.196.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.196.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.196.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.197.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.197.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.197.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.198.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.198.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.198.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.199.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.199.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.199.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.200.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.200.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.200.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.201.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.201.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.201.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.202.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.202.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.202.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.203.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.203.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.203.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.204.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.204.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.204.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.205.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.205.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.205.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.206.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.206.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.206.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.207.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.207.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.207.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.208.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.208.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.208.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.209.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.209.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.209.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.210.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.210.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.210.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.211.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.211.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.211.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.212.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.212.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.212.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.213.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.213.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.213.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.214.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.214.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.214.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.215.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.215.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.215.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.216.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.216.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.216.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.217.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.217.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.217.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.218.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.218.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.218.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.219.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.219.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.219.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:32 root] (omniquant.py 432): INFO model.layers.60.experts.220.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.220.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.220.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.221.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.221.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.221.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.222.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.222.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.222.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.223.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.223.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.223.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.224.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.224.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.224.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.225.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.225.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.225.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.226.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.226.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.226.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.227.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.227.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.227.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.228.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.228.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.228.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.229.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.229.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.229.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.230.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.230.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.230.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.231.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.231.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.231.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.232.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.232.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.232.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.233.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.233.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.233.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.234.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.234.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.234.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.235.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.235.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.235.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.236.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.236.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.236.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.237.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.237.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.237.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.238.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.238.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.238.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.239.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.239.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.239.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.240.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.240.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.240.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.241.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.241.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.241.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.242.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.242.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.242.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.243.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.243.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.243.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:33 root] (omniquant.py 432): INFO model.layers.60.experts.244.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.244.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.244.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.245.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.245.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.245.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.246.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.246.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.246.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.247.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.247.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.247.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.248.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.248.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.248.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.249.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.249.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.249.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.250.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.250.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.250.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.251.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.251.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.251.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.252.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.252.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.252.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.253.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.253.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.253.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.254.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.254.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.254.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.255.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.255.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.experts.255.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.shared_experts.mlp.gate_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.shared_experts.mlp.down_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 07:48:34 root] (omniquant.py 432): INFO model.layers.60.shared_experts.mlp.up_proj: QParams(128, [4], [1], 4, 32) +[2025-03-26 08:07:22 root] (omniquant.py 804): INFO layer 60 iter 0 loss:nan fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 08:25:36 root] (omniquant.py 804): INFO layer 60 iter 1 loss:nan fp_loss:0 norm:0.0 lr:5e-06 max GPU memory_allocated 74244.7373046875 +[2025-03-26 09:00:43 root] (main.py 408): INFO 226551.22566652298 diff --git a/model-00001-of-00073.safetensors b/model-00001-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..82d333ef80fdae5d07e6bc0dedcdd25c54006744 --- /dev/null +++ b/model-00001-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1be8cfc30e0e3aacbac727fb0e4d665207dc9813380bceeb9fd1248dfae98a00 +size 4995184776 diff --git a/model-00002-of-00073.safetensors b/model-00002-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a0447b04b1d9040e8d47dd942aff1c77cf83aed4 --- /dev/null +++ b/model-00002-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e1c320c5ce3a7af70717028a9da07cb621492c93a713879288d45b2a56fa7ae5 +size 4997775232 diff --git a/model-00003-of-00073.safetensors b/model-00003-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c5b3dd08ffaf09d6c889810cb5caf717e8eb0fd8 --- /dev/null +++ b/model-00003-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6f8fb977808d9e45ec2597803b2786131617e71331a36f367ebf14d84648e88a +size 4999329648 diff --git a/model-00004-of-00073.safetensors b/model-00004-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0404af82cb16101d5586c7fbd195de2ac9a2252a --- /dev/null +++ b/model-00004-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:464967b39525f68fc5bbc410830d3ee64e32346d61f516a5a08f32f049306e79 +size 4999329640 diff --git a/model-00005-of-00073.safetensors b/model-00005-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8f14df6522dd75b56ad05893b24c4cfa6f54d224 --- /dev/null +++ b/model-00005-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:247ef163860739e647d6680aec3193f5c18ffb168230b64b89d2a94f355b761b +size 4999350432 diff --git a/model-00006-of-00073.safetensors b/model-00006-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d116d044ddf0099dab0a61f5c763275457eef883 --- /dev/null +++ b/model-00006-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:10ee31c68578a2a2863d13e59a4e6a54387e7ba80359c268db43ac9c867c3ce8 +size 4999330368 diff --git a/model-00007-of-00073.safetensors b/model-00007-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ad0f54efa86800190937ab46441b9ff83f5ea2db --- /dev/null +++ b/model-00007-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:01aa1d2f4ac5f44626e5d0d66aee3cc5029d49e73ace4f65302ef28941b4d49e +size 4997775528 diff --git a/model-00008-of-00073.safetensors b/model-00008-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0f67c5d51737ba0185a79920ceee8e9e7c53440e --- /dev/null +++ b/model-00008-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d99b3b907c17d0d9deacd9dafc3097982bd74efb8234d39285ec1b2903d8adf1 +size 4999329640 diff --git a/model-00009-of-00073.safetensors b/model-00009-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b6df6fbb58af967d321ffbcd2f9e11823330c2d1 --- /dev/null +++ b/model-00009-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2b710b507d72715fa7a17514f0dfb437824df946fc612958ad077a4327c188e0 +size 4999350120 diff --git a/model-00010-of-00073.safetensors b/model-00010-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a74cea95079b61e54c6d9a25e38f6fa51768a4f3 --- /dev/null +++ b/model-00010-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b16fef9c17613abd7376c59747543c2869efa31b466961581c74176914c008de +size 4999331168 diff --git a/model-00011-of-00073.safetensors b/model-00011-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..18c3d395e5a897506c613737a248b5afe798ece2 --- /dev/null +++ b/model-00011-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:939c52abcb27dc3437015a367320dc2d6e35a62aabde4437fb18000e26d8dba9 +size 4999333528 diff --git a/model-00012-of-00073.safetensors b/model-00012-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1c471e89152ed1f6cbde6952fc3ab68efcfb4b89 --- /dev/null +++ b/model-00012-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7f3fa6cdb1f28a60a4b71388ff3c86e1cf698b0fe859081b0515550e35dc9869 +size 4999354120 diff --git a/model-00013-of-00073.safetensors b/model-00013-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..55938a480b719ff7acba9fe057ab125a7e8e34e5 --- /dev/null +++ b/model-00013-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ab2844dd9a4302ff0cade55268f088564583ba1aa964a31a4958945eec712d1f +size 4997778176 diff --git a/model-00014-of-00073.safetensors b/model-00014-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4671b0236c9028a54fe1c31c8cfa04ce3d9a67fc --- /dev/null +++ b/model-00014-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:81654da5d81ce62ffa39dc4cc09502b19ebfba3a04e28124b9fd3079892cff01 +size 4999332808 diff --git a/model-00015-of-00073.safetensors b/model-00015-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9fada9ede5838185a36de7f308437d2d45e4c1bb --- /dev/null +++ b/model-00015-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f7bb837482e3ee9622658cb8cff78308e5a7bc5cc6e806df0cb679991fad91f2 +size 4999332800 diff --git a/model-00016-of-00073.safetensors b/model-00016-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..409b172694eee4b6ad5cccfd6d70dbfe2ac53d82 --- /dev/null +++ b/model-00016-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8fa0411a4d46e24304245a49a5f07086344d4abf896b641754dabfef1e0a90b4 +size 4999353728 diff --git a/model-00017-of-00073.safetensors b/model-00017-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..57790a70c40ccf8d5e50e0390a8b803f4eca99a8 --- /dev/null +++ b/model-00017-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:046808a61c16d65fed158d2907d2355601c74ad8f15871809fa4b160c1379460 +size 4999333528 diff --git a/model-00018-of-00073.safetensors b/model-00018-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..de0d440a16f5c9f564ed3e1066d124fea2cdba12 --- /dev/null +++ b/model-00018-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c396baf86904a21f6914335def949ffa378d7274183f2aff26aec26c577c3d82 +size 4997778592 diff --git a/model-00019-of-00073.safetensors b/model-00019-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7febb291cec33cb1cf9bc0d13aed9c4ab90c5e17 --- /dev/null +++ b/model-00019-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:261789402ceba66afefb550b4de4c49120d5ffbd0a045da992658c22993da821 +size 4999332800 diff --git a/model-00020-of-00073.safetensors b/model-00020-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..97880de61b55fa68d6fb384cf4a194b2c68ec269 --- /dev/null +++ b/model-00020-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ae090c6e747d7ecc253481703c88fab5d2b41a3fc3c96a9cf7d229d2cf8b6f29 +size 4999353280 diff --git a/model-00021-of-00073.safetensors b/model-00021-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..39814d91886a1d621bfa72d4ca59c760a2f54526 --- /dev/null +++ b/model-00021-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:24dcf799d003411469eed92983cc4aa5ada66bc8d26a02803656276704142655 +size 4999332960 diff --git a/model-00022-of-00073.safetensors b/model-00022-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f201e2c3cc5347046d0022f8a5d52e7afb34fc5f --- /dev/null +++ b/model-00022-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e5d7e9c1b9b08e43e9ea47a4638a6814fc251c3eeaf8e7baae115873caa588bb +size 4999333528 diff --git a/model-00023-of-00073.safetensors b/model-00023-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e9971113a6d24ea9f09418232f278690c3fde342 --- /dev/null +++ b/model-00023-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:33c6619ada74bf25db6326075d2fd4808d32ec397702b8cdba78d5d80fa3cdeb +size 4997778888 diff --git a/model-00024-of-00073.safetensors b/model-00024-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1796fb8167b79fabb033ab0b13f8a4bd5303081a --- /dev/null +++ b/model-00024-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ce3ece860a1fc25edc4b70125161824538476d4105a5339cad01f3f52748ad9a +size 4999353280 diff --git a/model-00025-of-00073.safetensors b/model-00025-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9158405a8f69750dba0eb17ed0b2368e83f3613c --- /dev/null +++ b/model-00025-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c730571dc751944ec5e5490c15e15f24193f324f1d7149233c8ac835ee2d755f +size 4999332808 diff --git a/model-00026-of-00073.safetensors b/model-00026-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8ec03ce71611605ecc322ab79b904e807492f72d --- /dev/null +++ b/model-00026-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b7c01b937c06a9996670b4c013dd4ef4eaedd751cb35978a0ae5c40c2db8fad3 +size 4999332800 diff --git a/model-00027-of-00073.safetensors b/model-00027-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fc8c97b5ec62e7f6b62a595bf542dcd0d7fd11d7 --- /dev/null +++ b/model-00027-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:27606899f7697c7e4ea0c6757ecba940658b6b0088431fb2a256d335a6912d04 +size 4999353864 diff --git a/model-00028-of-00073.safetensors b/model-00028-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3511f4f6911b8b46ec9b9a87975667883bb9045f --- /dev/null +++ b/model-00028-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a04185dd86cbc3fc80f5bb74fbf7c58479f2759ab39c028b3f795499613b462c +size 4999333528 diff --git a/model-00029-of-00073.safetensors b/model-00029-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..491a8c854d4a86b3471cdaf8798ddd4a97bd4c32 --- /dev/null +++ b/model-00029-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:231d966ae8bca387d2dc74351a790c03cd0f247e48563066534a5eb8309403a6 +size 4997778456 diff --git a/model-00030-of-00073.safetensors b/model-00030-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e9b02071cd67efd02e849c2dc3bd9489c06510d7 --- /dev/null +++ b/model-00030-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d75e0d8a7e3b4e28baef4b977907c3dd7dea02d407cb810ba2799b283162009 +size 4999332800 diff --git a/model-00031-of-00073.safetensors b/model-00031-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f8b03f08bade952a4854ecf4aea46dcb0d6c1815 --- /dev/null +++ b/model-00031-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2b3705bc3f4952fcc5d44efcfafa74a4d53032fa5441a4bf19534540b152fa2e +size 4999353280 diff --git a/model-00032-of-00073.safetensors b/model-00032-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..39ba6cd2e5a50d2e30167efd95becf8b91258e68 --- /dev/null +++ b/model-00032-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c473f6c6b031d1aa015573996a10ba2fee658b60b0b2a05e1faa480dd246dd3d +size 4999333096 diff --git a/model-00033-of-00073.safetensors b/model-00033-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..dcc0f59da027259fba893f4b3cd137f49bfb7cfb --- /dev/null +++ b/model-00033-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:370fffd610f8621a2d0ca9c41364cd13c3af2bc20fd70d9cd606c1d97733f7c1 +size 4999333528 diff --git a/model-00034-of-00073.safetensors b/model-00034-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..385179f37e2cc945d03e80d92a3850bb81f19291 --- /dev/null +++ b/model-00034-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5a35fb71d1cadcd9a91ce7e0905c02d5d85466efaab9c14b54e4200bd0ff3f88 +size 4997778752 diff --git a/model-00035-of-00073.safetensors b/model-00035-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..cc0a192b478091a06baf69d6ec96806fa969a7c8 --- /dev/null +++ b/model-00035-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f51d187d780f325b5514cc1038074454e92a1700607df5d71bb137cd14d023b +size 4999353280 diff --git a/model-00036-of-00073.safetensors b/model-00036-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..eda4c0d76ca31cd33ecd41ecc9d2e6bb5f1e0a27 --- /dev/null +++ b/model-00036-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4e4f3b0e2b9b30890e58e62d387c986c00bf8042aad782707a08559d87fb1456 +size 4999332808 diff --git a/model-00037-of-00073.safetensors b/model-00037-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4d8b3aef841825ed4954b93e2d231c816b66349e --- /dev/null +++ b/model-00037-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2951e359e59d76c374276d2cb831ece8f164a5bde4840f294142a3f30c41eb66 +size 4999332800 diff --git a/model-00038-of-00073.safetensors b/model-00038-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9f19c55d1a08424dc78f5ceea0e3a8fa3de90931 --- /dev/null +++ b/model-00038-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a9f416cec45088bd621d51b56bb7c8d96ede71b3aa5f16bf477c9e2a94d82549 +size 4999354000 diff --git a/model-00039-of-00073.safetensors b/model-00039-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..21942682a7243dc9077997857dba5e59d3317bd6 --- /dev/null +++ b/model-00039-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6e58cc01a09abd3cd3830ab5e59a4f479d7961f752b4a5d864575def1e099168 +size 4999333624 diff --git a/model-00040-of-00073.safetensors b/model-00040-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..64fb64760314a607a21cf5e0b1152b6a22170292 --- /dev/null +++ b/model-00040-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:feea2082444cce4aa12121712df94afb5f8062a8119db0e200e3f54787862649 +size 4997778224 diff --git a/model-00041-of-00073.safetensors b/model-00041-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..061cd6e75ae9adf26136acc7973e4c01894d718c --- /dev/null +++ b/model-00041-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:dbec8c76233f0f75b59f26ed8c5f625dee86ece89c921ead3a9a3c7e89c6e609 +size 4999332800 diff --git a/model-00042-of-00073.safetensors b/model-00042-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..89ff1803adbc0f3bc0572c6e91bfa361fdd90713 --- /dev/null +++ b/model-00042-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c1ec2c7fedec37a9fc9f901b89a996730246836f420c8eec170c008e5437bed8 +size 4999353280 diff --git a/model-00043-of-00073.safetensors b/model-00043-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..47924ad2ab69f883e61cab7bb26d1ed7b107e1e8 --- /dev/null +++ b/model-00043-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1bc7a0194e2b5ab1491f1748e6ab7f0c08c85a13955c301be1d4576a92df5922 +size 4999333232 diff --git a/model-00044-of-00073.safetensors b/model-00044-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c019055071d9e065ac453ee1e107a72dc571ae4d --- /dev/null +++ b/model-00044-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d81a5b8530f98ce4771c3af3875fbf0b1a7008e7dc27bd98699b804c2b6a93b8 +size 4999333528 diff --git a/model-00045-of-00073.safetensors b/model-00045-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..992083d98b8600c5330a3855b18cb938c9f954ea --- /dev/null +++ b/model-00045-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a93a97280e4c8003d3b651c6189811e7c78bf7bdd554935550e0cc6e9b3a012f +size 4997778616 diff --git a/model-00046-of-00073.safetensors b/model-00046-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..619e5c5cce0b8d1a677ecbc16f57498453a00b11 --- /dev/null +++ b/model-00046-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8a48d0a1584f31e8f69a0f077982749dda0d2a253833c0187bfc9639b056ba5b +size 4999353280 diff --git a/model-00047-of-00073.safetensors b/model-00047-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..94083374ec53e34a5fff038a55650ef4596ae18d --- /dev/null +++ b/model-00047-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:47c62a1ff4631713d1d4e6565950457c2c0bb9a38e1f978ce2b95c1333fc7a49 +size 4999332808 diff --git a/model-00048-of-00073.safetensors b/model-00048-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..08825be1c0fd15546888b6ebee11198589207902 --- /dev/null +++ b/model-00048-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:770a2850329c6cdee516dc3c72eae371286cc26f9afe226522df14bd35bcbeb1 +size 4999332928 diff --git a/model-00049-of-00073.safetensors b/model-00049-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..cc7ac5ef0c760a4b5472d1651bcdb5a21d0486c5 --- /dev/null +++ b/model-00049-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:043bfd6cb4efecf4edd19189d1bfbcb2ee2e9acf342e22bfe6c47cf578247b31 +size 4999354008 diff --git a/model-00050-of-00073.safetensors b/model-00050-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..51a78d820884b092e9ca5964e81ad0fd094a270e --- /dev/null +++ b/model-00050-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4cbe666a5f918642020ccd4f9664410078234d75bb36c15575ed5f33d37518e7 +size 4997749984 diff --git a/model-00051-of-00073.safetensors b/model-00051-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c1957b80adb9caf6b2e285dba22daa51b8052f11 --- /dev/null +++ b/model-00051-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:05d2985200faeb59170dafb2026cfee2fab6d85bf2c4b90eae79e55b8c54f305 +size 4999361728 diff --git a/model-00052-of-00073.safetensors b/model-00052-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b07ebb23090533ef6464d1d57daf690fb10234e6 --- /dev/null +++ b/model-00052-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:32573bde8e005f872f1c6299cbc9c27f05d39e1bae0d751097f1176806b21ae5 +size 4999332800 diff --git a/model-00053-of-00073.safetensors b/model-00053-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..76bc42704b83d3336d31dfc3abf2c2d4b105dcdf --- /dev/null +++ b/model-00053-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6e9175aee52e8078be8f58016f8972eaaedb0b21799418e8680e4b14d0aaa458 +size 4999353280 diff --git a/model-00054-of-00073.safetensors b/model-00054-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..db6a39eadf97d1b7c3491e92a4d3c15f4376d35f --- /dev/null +++ b/model-00054-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e241290f67a78edddcc081ac9703e93c70b79943c59bd000f3c0435fb6f9a1f5 +size 4999333360 diff --git a/model-00055-of-00073.safetensors b/model-00055-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e1905a5e3a8271f45462de1ded78d2187cdccccb --- /dev/null +++ b/model-00055-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0316f0403d11dbb6b62092414ea7863fef25dd861c88492ecd2fa963a94f0574 +size 4999333528 diff --git a/model-00056-of-00073.safetensors b/model-00056-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8bebf27c2ea00887efbf267aa74d23390941e961 --- /dev/null +++ b/model-00056-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:336e36b0cd82ae3edc7625d6706aaf45ccb5747125cddf0d2881fd7b52dbb906 +size 4997778480 diff --git a/model-00057-of-00073.safetensors b/model-00057-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ef6fe7bdabf50b9d8759ed8bb6cbdc48a4497e08 --- /dev/null +++ b/model-00057-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b896cf54483bcc1d74aea254e126b1350708acf3c74c826696ea930d87c3aab5 +size 4999353280 diff --git a/model-00058-of-00073.safetensors b/model-00058-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5773dd1e00d97cec7dfb8b4045089358307cf190 --- /dev/null +++ b/model-00058-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ef0a3d20307b191a8ee9aabc300be691f40232f6e2d076f73663708e558318e3 +size 4999332808 diff --git a/model-00059-of-00073.safetensors b/model-00059-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f833a4d7f53809df0bbb29fcd6a36629180b6482 --- /dev/null +++ b/model-00059-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a5c449411f2a5dfd73f083a7bddd8aa468d92cfdbbc0c0c3af8563873ef4af61 +size 4999333064 diff --git a/model-00060-of-00073.safetensors b/model-00060-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ef3ad58afdef6d8e8b32f57e459751480954873c --- /dev/null +++ b/model-00060-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:80b5d87be7d6d57377712153c33e1319da7f2c320a6dc8c31295af4222e04330 +size 4999354008 diff --git a/model-00061-of-00073.safetensors b/model-00061-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ac8138a68a41af757b1a029217b3c7e9150f8a56 --- /dev/null +++ b/model-00061-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:13681b3efc24b7fe0aa40081585e7cc3e0745aef4c4e74b816b25f65264583de +size 4997778776 diff --git a/model-00062-of-00073.safetensors b/model-00062-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a6bd26bf40057079e779e8952c8cd1578ede09d7 --- /dev/null +++ b/model-00062-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8d7a03b351a8d65e05f04d0b9fcff9c12600bad14ad065bf5b3bc2b7b2ceb243 +size 4999332808 diff --git a/model-00063-of-00073.safetensors b/model-00063-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9be9ef595ed503e70e0bada7c20b7668f6295d02 --- /dev/null +++ b/model-00063-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a5b18d73dc6ea27486136ee6806d984ec3d69d344adab42ae5fd0efbc2cf228b +size 4999332800 diff --git a/model-00064-of-00073.safetensors b/model-00064-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a5754928b9224e8b042d269c714422755446677c --- /dev/null +++ b/model-00064-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2020a0bb634073446c949a137023f5360852f67f75ce0117352b510d073e88f3 +size 4999353280 diff --git a/model-00065-of-00073.safetensors b/model-00065-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..27a3095343b8bfe7c1cab2a064e9bc05a2a1bf6a --- /dev/null +++ b/model-00065-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cf24eba858a879b7e1393eedd0e823cc856ca53aa41323a92bb231c8bdd71008 +size 4999333496 diff --git a/model-00066-of-00073.safetensors b/model-00066-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..33fc19dbffc091397edefb3d7e8eee887f640a2d --- /dev/null +++ b/model-00066-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:14ab2fa1d8f46b2f05480ac0d7a1c89a9a5987fd55b6fef47589e2f2ad6fb007 +size 4999333600 diff --git a/model-00067-of-00073.safetensors b/model-00067-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..eafeecce14a24e8e6521531462e582c2b7fa9dbc --- /dev/null +++ b/model-00067-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:51434965491f1c9bb722ddefbde5ee409e0415f3476f54430202d554efccc20d +size 4997778280 diff --git a/model-00068-of-00073.safetensors b/model-00068-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a44d184372da00df41b3dbef097a8d325020a142 --- /dev/null +++ b/model-00068-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3c90120352f105d84462ce85ea7607474dc37c0dd0991d66d5bb447443a34134 +size 4999353280 diff --git a/model-00069-of-00073.safetensors b/model-00069-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..10019c2b890ca3a2603ff8e7f60adbd8481062c3 --- /dev/null +++ b/model-00069-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c28175ac4b3d623cbdc0c4c8d733bef8083a6581c0d8b0adf46cf8ac9f652e7f +size 4999332808 diff --git a/model-00070-of-00073.safetensors b/model-00070-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1d232aa3006c12ce771bd4cbff552ca4f3129121 --- /dev/null +++ b/model-00070-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ab5ba3200c6d452e22bdb24875ac47847db55d0c6ab91e1ee2a0ad955bfeeba7 +size 4999333200 diff --git a/model-00071-of-00073.safetensors b/model-00071-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a7cc13929fb8501fc15b219c9248a3394b98b40c --- /dev/null +++ b/model-00071-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:07aed8aae092fec1d6138a74419ef58cac176e8d9dec11eda307b0645c226a6d +size 4999354008 diff --git a/model-00072-of-00073.safetensors b/model-00072-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..baa0aa03a3570be2c73171e32fc02a9d673da7a1 --- /dev/null +++ b/model-00072-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f4315173f0f2ec3dd5ec9b278a189d923e359991f8dee11151831f1b9ec034a +size 4997778640 diff --git a/model-00073-of-00073.safetensors b/model-00073-of-00073.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..815e6497abfa65c99377d7018e83fe906ac635fc --- /dev/null +++ b/model-00073-of-00073.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1c0d4941797482da09fdc68792cb5112b4737d1092be1eaa98c6d5e0658e7dfb +size 4156227192 diff --git a/quant_strategy.json b/quant_strategy.json index f5636e371a0f463f899a2936f1ad60c957c4f648..1b02d4e67d3df1a97799f48183127218caff3d8b 100644 --- a/quant_strategy.json +++ b/quant_strategy.json @@ -1,7 +1,7 @@ { "measurement": { "model.layers.0": { - "accuracy": 0.9817549798171967, + "accuracy": 0.9992622286081314, "total_bits": 2379451392.0, "q_a_proj": { "group_size": { @@ -109,7 +109,7 @@ } }, "model.layers.1": { - "accuracy": 0.9940149370813742, + "accuracy": 0.9999234825372696, "total_bits": 2379451392.0, "q_a_proj": { "group_size": { @@ -217,7 +217,7 @@ } }, "model.layers.2": { - "accuracy": 0.9847580479981843, + "accuracy": 0.9994927197694778, "total_bits": 2379451392.0, "q_a_proj": { "group_size": { @@ -325,7 +325,7 @@ } }, "model.layers.3": { - "accuracy": 0.9904986454348546, + "accuracy": 0.9998161196708679, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -477,7 +477,7 @@ } }, "model.layers.4": { - "accuracy": 0.9895119229331613, + "accuracy": 0.9997029602527618, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -629,7 +629,7 @@ } }, "model.layers.5": { - "accuracy": 0.9867525280278642, + "accuracy": 0.9995767772197723, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -781,7 +781,7 @@ } }, "model.layers.6": { - "accuracy": 0.9837914108939003, + "accuracy": 0.9993880540132523, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -933,7 +933,7 @@ } }, "model.layers.7": { - "accuracy": 0.982880576106254, + "accuracy": 0.9992910772562027, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -1085,7 +1085,7 @@ } }, "model.layers.8": { - "accuracy": 0.9805211594211869, + "accuracy": 0.9992452412843704, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -1237,7 +1237,7 @@ } }, "model.layers.9": { - "accuracy": 0.9770258169737644, + "accuracy": 0.9979685992002487, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -1389,7 +1389,7 @@ } }, "model.layers.10": { - "accuracy": 0.9744746365468018, + "accuracy": 0.9981853365898132, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -1541,7 +1541,7 @@ } }, "model.layers.11": { - "accuracy": 0.9777291420614347, + "accuracy": 0.9984678477048874, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -1693,7 +1693,7 @@ } }, "model.layers.12": { - "accuracy": 0.9809753063018434, + "accuracy": 0.9987831711769104, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -1845,7 +1845,7 @@ } }, "model.layers.13": { - "accuracy": 0.9880705887626391, + "accuracy": 0.998877227306366, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -1997,7 +1997,7 @@ } }, "model.layers.14": { - "accuracy": 0.9904032560880296, + "accuracy": 0.9988386332988739, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -2149,7 +2149,7 @@ } }, "model.layers.15": { - "accuracy": 0.9941651911940426, + "accuracy": 0.9986088275909424, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -2301,7 +2301,7 @@ } }, "model.layers.16": { - "accuracy": 0.9934399027988547, + "accuracy": 0.9984094947576523, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -2453,7 +2453,7 @@ } }, "model.layers.17": { - "accuracy": 0.9944614085834473, + "accuracy": 0.998946025967598, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -2605,7 +2605,7 @@ } }, "model.layers.18": { - "accuracy": 0.9921934259618865, + "accuracy": 0.9983946084976196, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -2757,7 +2757,7 @@ } }, "model.layers.19": { - "accuracy": 0.9912811828980921, + "accuracy": 0.9979821741580963, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -2909,7 +2909,7 @@ } }, "model.layers.20": { - "accuracy": 0.9900320051528979, + "accuracy": 0.9981323629617691, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -3061,7 +3061,7 @@ } }, "model.layers.21": { - "accuracy": 0.9901153031969443, + "accuracy": 0.9983349591493607, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -3213,7 +3213,7 @@ } }, "model.layers.22": { - "accuracy": 0.9864137352560647, + "accuracy": 0.9976826012134552, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -3365,7 +3365,7 @@ } }, "model.layers.23": { - "accuracy": 0.9857180436956696, + "accuracy": 0.997639924287796, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -3517,7 +3517,7 @@ } }, "model.layers.24": { - "accuracy": 0.9841050353425089, + "accuracy": 0.9975894391536713, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -3669,7 +3669,7 @@ } }, "model.layers.25": { - "accuracy": 0.9832171808811836, + "accuracy": 0.9978430569171906, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -3821,7 +3821,7 @@ } }, "model.layers.26": { - "accuracy": 0.9834447980683763, + "accuracy": 0.9978500455617905, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -3973,7 +3973,7 @@ } }, "model.layers.27": { - "accuracy": 0.9812710456317291, + "accuracy": 0.9976160824298859, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -4125,7 +4125,7 @@ } }, "model.layers.28": { - "accuracy": 0.9811981529928744, + "accuracy": 0.9977917373180389, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -4277,7 +4277,7 @@ } }, "model.layers.29": { - "accuracy": 0.9799808135721833, + "accuracy": 0.9977639466524124, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -4429,7 +4429,7 @@ } }, "model.layers.30": { - "accuracy": 0.9797754119499587, + "accuracy": 0.9980919808149338, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -4581,7 +4581,7 @@ } }, "model.layers.31": { - "accuracy": 0.9786722971475683, + "accuracy": 0.9979510605335236, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -4733,7 +4733,7 @@ } }, "model.layers.32": { - "accuracy": 0.9808280722354539, + "accuracy": 0.9983908236026764, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -4885,7 +4885,7 @@ } }, "model.layers.33": { - "accuracy": 0.9827347988612019, + "accuracy": 0.9987252056598663, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -5037,7 +5037,7 @@ } }, "model.layers.34": { - "accuracy": 0.9827491079631727, + "accuracy": 0.998604416847229, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -5189,7 +5189,7 @@ } }, "model.layers.35": { - "accuracy": 0.9832003622432239, + "accuracy": 0.9987341612577438, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -5341,7 +5341,7 @@ } }, "model.layers.36": { - "accuracy": 0.9827660944429226, + "accuracy": 0.9987073093652725, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -5493,7 +5493,7 @@ } }, "model.layers.37": { - "accuracy": 0.98276356497081, + "accuracy": 0.9988316744565964, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -5645,7 +5645,7 @@ } }, "model.layers.38": { - "accuracy": 0.9846183339541312, + "accuracy": 0.9990833252668381, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -5797,7 +5797,7 @@ } }, "model.layers.39": { - "accuracy": 0.9832595270127058, + "accuracy": 0.9989469796419144, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -5949,7 +5949,7 @@ } }, "model.layers.40": { - "accuracy": 0.9807395663810894, + "accuracy": 0.9987863302230835, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -6101,7 +6101,7 @@ } }, "model.layers.41": { - "accuracy": 0.9818256291910075, + "accuracy": 0.9989507049322128, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -6253,7 +6253,7 @@ } }, "model.layers.42": { - "accuracy": 0.9831473883823492, + "accuracy": 0.9990694224834442, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -6405,7 +6405,7 @@ } }, "model.layers.43": { - "accuracy": 0.9827884282858577, + "accuracy": 0.9991038888692856, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -6557,7 +6557,7 @@ } }, "model.layers.44": { - "accuracy": 0.9825569812674075, + "accuracy": 0.9991481602191925, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -6709,7 +6709,7 @@ } }, "model.layers.45": { - "accuracy": 0.9813435177784413, + "accuracy": 0.9990146160125732, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -6861,7 +6861,7 @@ } }, "model.layers.46": { - "accuracy": 0.9808638134854846, + "accuracy": 0.9990020841360092, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -7013,7 +7013,7 @@ } }, "model.layers.47": { - "accuracy": 0.9815339095075615, + "accuracy": 0.9991412907838821, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -7165,7 +7165,7 @@ } }, "model.layers.48": { - "accuracy": 0.9809085759334266, + "accuracy": 0.998986005783081, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -7317,7 +7317,7 @@ } }, "model.layers.49": { - "accuracy": 0.9807316974038258, + "accuracy": 0.9991100281476974, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -7469,7 +7469,7 @@ } }, "model.layers.50": { - "accuracy": 0.9797588220098987, + "accuracy": 0.9989775270223618, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -7621,7 +7621,7 @@ } }, "model.layers.51": { - "accuracy": 0.9814481334178708, + "accuracy": 0.9991746246814728, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -7773,7 +7773,7 @@ } }, "model.layers.52": { - "accuracy": 0.9798667979193851, + "accuracy": 0.9990788698196411, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -7925,7 +7925,7 @@ } }, "model.layers.53": { - "accuracy": 0.9805318327853456, + "accuracy": 0.9991910457611084, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -8077,7 +8077,7 @@ } }, "model.layers.54": { - "accuracy": 0.9779171335394494, + "accuracy": 0.9987901151180267, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -8229,7 +8229,7 @@ } }, "model.layers.55": { - "accuracy": 0.9781052877078764, + "accuracy": 0.9989645630121231, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -8381,7 +8381,7 @@ } }, "model.layers.56": { - "accuracy": 0.9785591780673712, + "accuracy": 0.9990920573472977, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -8533,7 +8533,7 @@ } }, "model.layers.57": { - "accuracy": 0.9786476686713286, + "accuracy": 0.9991132915019989, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -8685,7 +8685,7 @@ } }, "model.layers.58": { - "accuracy": 0.9801497724256478, + "accuracy": 0.9992498308420181, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -8837,7 +8837,7 @@ } }, "model.layers.59": { - "accuracy": 0.9751776319462806, + "accuracy": 0.9988341629505157, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -8989,7 +8989,7 @@ } }, "model.layers.60": { - "accuracy": 1e-06, + "accuracy": 0.9963512718677521, "total_bits": 46949960704.0, "q_a_proj": { "group_size": { @@ -9141,4 +9141,4 @@ } } } -} +} \ No newline at end of file